O New Zealan
PRACTICAL FEBRUARY 1985 -£1-00
ROBOTICS -MICROS - ELECTRONICS - INTERFACING
MODULAR AUDIO
POWER SYSTEM
SPECTRUM
DAC/ADC BOARD
SIGNAL
GENERATOR...
F-V CONVERTER
MICROWRITING —
The Principle
and the Product
ROBOTICS
MICROS
| ELECTRONICS
INTERFACING at
VOLUME 21 NG2 FEBRUARY 1985
CONSTRUCTIONAL PROJECTS
MODULAR AUDIO POWER SYSTEM—Part 1 by WV. Tooley BA and D. Whitfield MA MSc CEng MIEE.. 10
Main power amp module
SPECTRUM DAC/ADC BOARD by f.A. Penfold .. = af ne we 9 me ~ 15
Interface board for control applications
MICROPROCESSOR CONTROLLED D.C. MOTOR DRIVERS by 7om Gaskell BA(Hons) CEng MIEE.. 31
Enables analogue driving of d.c. motors
SIGNAL GENERATOR & F-V CONVERTER by John M. H. Becker gi bs sna ia .. 34
Quality test instrument
NEPTUNE AND MENTOR ROBOTS by Richard Becker and Tim Orr ‘3 vs ee te .. 49
Part Six: Commissioning and testing of Neptune
MONO/STEREO CHORUS & FLANGER by John M.H. Becker .. 4% sts ou ats .. 589
Part Two: Construction and setting up
GENERAL FEATURES
MICROWRITER by Jom Gaskell BA(Hons) CEng MIEE ssa i is or - a si 22
An ingenious six-key alternative to the QWERTY keyboard
SEMICONDUCTOR CIRCUITS by Tom Gaskell BA(Hons) CEng MIEE wt i = ee .. 28
Power Op-Amps (TCA 365 and TCA 2365)
SPACEWATCH by Dr. Patrick Moore OBE... me sig zs oa a i ie .. 40
INGENUITY UNLIMITED At . 42
Readers’ circuit ideas
SEQUENTIAL LOGIC TECHNIQUES by ™. Tooley BA and D. Whitfield MAMSc CEng MIEE .. 6S
Part Five: Data multiplexers
NEWS & COMMENT
EDITORIAL = s 7 BAZAAR . . 18,48 MICROBUS ae -. 6§
NEWS & MARKET INDUSTRY NOTEBOOK.. 21 VERNONTRENT .. »a G7
PLACE... ee is 8 LEADINGEDGE .. .. 25 P.C.B.SERVICE .. -. 68
RRACTICAL PEERUARY AR 1007
ORCS | NCE METRIC INT ACN
THIS MONTH'S
COVER...
Our cover photograph
shows silicon wafers in a
furnace during the produc-
tion of integrated circuits.
Photograph courtesy of
Nationa! Semiconductor.
OUR MARCH ISSUE WILL BE ON SALE FRIDAY, FEBRUARY 1st, 1985 (see page 47)
© IPC Magazines Limited 1985. Copyright in all drawings, photographs and articles published in PRACTICAL ELECTRONICS is
! fully protected, and reproduction or imitations in whole or part are expressly forbidden. All reasonable precautions are taken by
; PRACTICAL ELECTRONICS to ensure that the advice and data given to readers are reliable. We cannot, however, guarantee it,
and we cannot accept legal responsibility for it. Prices quoted are those current as we go to press.
Practical Electronics February 1985
| E VOLUME 21 N°G2 FEBRUARY 1985
SAFETY
LECTRONICS has done much to
benefit our way of life and stan-
dard of living in all areas from enter-
tainment to safety at sea and in the
air. Of course it has also enabled
development of more sophisticated
weapons and defence systems but
that is another story. Our explora-
tion of space is totally dependent on
electronics and navigation about our
own planet is also now based
mainly on high technology.
What a pity then that the modes
of transport we ail use every day
have not benefited more from the
introduction of electronics to aid
safety. The car you drive may have a
computer to show fuel consump-
tion, it may have a talking dash
pane! or even an engine manage-
ment computer, but have the elec-
tronics been used to improve
safety? How many vehicles are
fitted with an anti-locking braking
system? How often do you see
vehicles skidding even on dry roads?
How often do the back wheels of
unladen lorries lock up when they
stop? How many motorcyclists
come off in the wet when braking or
skid into the back of the car in front?
Admittedly many of the skids that
do occur result in no damage or in-
jury but of course some do. Surely it
is better to make vehicles safer with
an electronically controlled failsafe
braking system than to get them
talking to you? This is one area
where the amateur in electronics
can do little himself. We would not
encourage readers to modify any
vehicle braking system, so we can-
not fit a system to help ourselves.
The sad thing is that the
technology and mechanics to per-
form the necessary tasks has been
around for some years. Perhaps the
manufacturers feel we wil! not pay
for the extra safety; maybe they do
not feel it is necessary? The next
time you see a minor skid that could
have been dangerous, a motor-
cyclist fall off, or a lorry stopping
slightly sideways just think about
what could have gone badly wrong
and see if you feel anti-skid braking
would be worth another couple of
hundred pounds on the already in-
flated price of a new vehicle in the
UK,
LEGISLATION
Maybe you will even think that
legislation would be a good thing,
even if it might not save as many
lives as compulsory seat belt
wearing!
Incidentally, the motorcyclist |
saw come off this morning was
shaken but not badly injured,
although his bike was probably a
write-off and the car he ran into
badly damaged, Think about it if you
buy a new vehicle! The extra cost
could be worth the time, trouble and
heartache alone.
BACK NUMBERS
and BINDERS...
=. Copies of most of our recent
issues are available from: Post
Sales Department (Practical Elec-
tronics), IPC Magazines Ltd.,
Lavington House, 25 Lavington
Street, London SE1 OPF, at £1
each including Inland/Overseas
p&p. Please state month and
year of issue required,
Binders for PE are available
from the same address as back
numbers at £5.50 each to UK or
overseas addresses, including
postage, packing and VAT
NICAD CHARGER
Practical Electronics February t985
Editor Mike Kenward
Secretary Pauline Mitchell
Editorial Tel: Poole (0202) 671191
Advertisement Manager
David Tilleard 01-261 6676
Secretary
Christine Pocknell 01-261 6676
Classified Supervisor
Barbara Blake 01-261 5897
Ad. Make-up/Copy
Brian Lamb 01-261 6601
Queries and letters concerning
advertisements fo:
Practical Electronics Advertisements,
King’s Reach Tower,
Stamford Street, London SE1 9LS
Telex: 915748 MAGDIV-G
Letters and Queries
We are unable to offer any advice on the
use or purchase of commercial equipment
or the incorporation or modification of
designs published in PE. All letters requiring
a reply should be accompanied by a stam-
ped addressed envelope, or addressed en-
velope and international reply coupons, and
each letter should relate to one published
project only.
Components are usually available from
advertisers: where we anticipate difficulties
a source will be suggested.
Old Projects
We advise readers to check that all parts
are still available before commencing any
project in a back-dated issue, as we cannot
guarantee the indefinite availability of com-
ponents used.
Technical and editorial queries and /etters to.
Practical Electronics Editorial,
Westover House,
West Quay Road, Poole,
Dorset BH15 1JG
SUBSCRIPTIONS
Copies of Practical Electronics are available
by post, inland for £13, overseas for £14
per 12 issues, from: Practical Electronics,
Subsceription Department, |PC Magazines
Ltd., Room 2816, King’s Reach Tower,
Stamford Street, London SE1 9LS. Che-
ques, postal orders and international money
orders should be made payable to IPC
Magazines Limited. Payment for subscrip-
tions can also be made using a credit card.
Phone:
Editorial Poole (0202) 671191
We regret that lengthy technical enquiries
cannot be answered over the telephone.
tmentioned are available
normal retail outlets,
otherwise specified.
correct at time of going
HIGH COST INSULATION
AMRESUGSS o
Most constructors will be painfully aware of the annoying shrink-back
properties of insulation, encountered when soldering wires into place.
Over the years manufacturers have developed heat-proof insulation
materials for specialised cable applications which eventually filter
through industry to the home-constructor—and very welcome they are,
too. It.may surprise you to know, however, just how far, and to what ex-
pense, manufacturers will go in order to optimise the insulating properties
of the materials they use.
B.I.C.C. for instance has just completed
the installation of a new electron beam ac-
celerator plant at its Electronic Cables fac-
tory in Cheshire, the cost? A staggering
£2-5 miltion. The facility is considered to be
the most sophisticated and versatile of its
kind in the Western world. The accelerator
produces high velocity electrons which
have sufficient energy to penetrate the
cable insulation.
Once inside a polymeric insulation, the
electrons initiate chemical reactions which
lead to the formation of chemical bonds or
crosslinks. Increasing the number of
crosslinks leads to eventual formation of a
three-dimensional network which substan-
tially enhances the physical properties of
the insulant.
The most obvious effect of crosslinking is
that the material loses its thermoplastic
characteristics and becomes a non-melting
with a better balance of
mechanical properties at both high and low
temperatures; chemical resistance is also
enhanced.
The whole facility is enclosed in 1500
tonnes of concrete for personnel protection
during plant operation. The picture shows
the plants computer contro! room.
thermoset
TN TELEVISION/ MONITOR
The latest in the TX range from Ferguson is
a 14 inch monitor/colour television. It will
offer those who can afford a second or even
third set a very flexible visual display tool.
The MCO1 has separate RGB, composite
video and aerial inputs enabling the user to
get the best possible display from broad-~-
cast TV, video recorders, teletext and home
computers.
Perhaps the most interesting of these
options is the ability to directly connect
a home computer without the
modulation/demadulation problems that
occur when using a standard TV set. |t must
be borne in mind that not all currently
available home computers have a direct
video output. The machines without
this facility have on-board
modulation/demodulation and weré so
designed for use with a visual display
medium that most people already
possessed—a standard TV set.
The provision of separate RGB, com-
posite video and aerial! sockets also allows
the home computer, video recorder or
game and TV aerial to be connected
simultaneously; the set senses the signal
selected and switches to it automatically.
A range of special connector leads is
available to cover the different home com-
puter options. The set is manufactured in
the UK at the company’s Gosport plant. It is
expected to retail at circa £230.
PATENTLY
OBVIOUS
All too often the most worrying aspect for the
creator of an original design is, how to protect
that idea from those who would copy and ex-
ploit it for their own gain. This has been the
‘case since the first inventor brought forth a
brainchild, only to stand by helplessly as
someone else marketed his idea and made
a fortune. The laws governing Patents,
Trademarks, Designs and Copyright are com-
plex indeed, without guidance the layman may
be forgiven for getting confused. Laurence
Shaw’s recently reprinted guide can be of
great help to inventors and innovators alike.
The Practical Guide for people with a new
idea is a book which explains in clear
language how to protect a new idea, product
or scheme and exploit it to the full. Market
research, approaching a manufacturer, telling
the world about an idea without losing your
rights and patenting an invention are all
covered together with secret patents.
This publication is available from
booksellers at £5.50, or by mail order at £5.95
from The Patent Eye, George House, George
Road, Edgbaston, Birmingham B15 IPG.
(021 454 2165).
MAGIC
LANTERN
Question: If you are exposed to radiation do
you glow in the dark? Answer: Of course not,
Not unless you are first coated with a
phosphor of some kind. It is a useful fact that
beta particles from a radioactive source will,
when they strike a phosphor such as zinc sul-
phide, cause light to be emitted from it. Bat-
telle’s: Pacific’ Northwest Laboratories are
testing a novel application of this
phenomenon. Scientists are evaluating a por-
table runway lamp for setting up landing
strips in out-of-the-way places, or during
emergencies in which the electricity supply is
lost,
The lamp comprises a glass tube, its inside
surface coated with a phosphor, and which is
filled with tritium gas, the radioactive isotope
of hydrogen. The lamp can not be turned off, it
simply continues to glow for the twelve years
half-life of the gas. Keeping the glass clean is
the only maintenance operation required dur-
ing that time. The quantity of radioactive
material used is so minute that it is harmless
even if the glass breaks, it is claimed.
During field tests in Alaska pilots reported
that they perceived light from the
radioluminescent lamps differently from that
of conventional light, and human response
now needs to be assessed to find out how
useful these lamps may be.
1
= Practi¢al Electronics. February 1985
MARKEZ PI.
BTs rumble
machine
See Oe a
British
‘TELECOM, °
it's new from British Telecom,
For paging far-off staff.
A little pocket thing,
That could well cause a laugh.
You see instead of ‘bleeping’,
it’s been made to ‘vibrate’.
So you're the only one that knows,
HQ and you have got a date.
The waveforms coming through the air,
Will go right through your pants.
And trigger-off this rumble-box,
Like a herd of elephants.
So if you're in a meeting,
Friends might still get the rise.
When they notice that your eyeballs,
Are looking like mince pies.
HEADS-U-WIN
Ensuring peak response and high-quality
reproduction, Electrofube’s Video Tape
Head Cleaner is a safety solvent designed
for use on all magnetic tape heads.
The cleaner loosens and removes ac-
cumulated deposits of dirt and tape oxide
and dries quickly without leaving any
residues on the tape. The cleaner is non-
flammable, and non-conductive, it will not
damage plastics or rubber.
The sgivent comes in handy 110 gram
aerosols and jis conveniently applied by
spraying directly onto the heads and
mechanisms. {in addition, the cleaner is
ideal for spraying onto cleaning tapes and
other tape cleaning devices, such as cotton
buds or felt and chamois leather sticks.
Available on its own at circa £1.20, or
with 25 extra long cotton buds at circa
£1.60 from electrical retail outlets. ;
BBC's 00 Be
Following a tongue-in-cheek comment from ©
Mike Cook, the Technical Editor of Micro
User, several hundred BBC micro owners
recently returned their machines to their
respective dealers, in the fear that they were
about to detonate.
The unfortunate comment was printed in
the magazine's problem page as part of a reply
fo a reader's query regarding an ‘error
message’. Mr Cook, believing himself to be the
subject of a “wind-up”, answered in kind.
“Take your computer immediately to the
dealer as this error message indicates that it is
about to explode.”
The manufacturers, Acorn Computers,
were not amused, neither was the middle-aged
housewife who reportedly surrounded her
machine with a bucket of water,
POINTS
ARISING...
RING MODULATOR Ses
Ma” To
December, 1984 ts
Alterations to this project must be made as
follows:
In Fig. 9 the component marked C35
should be marked R35.
The capacitor C21 should have its +ve
terminal connected to R10. :
In Fig, 10 the unmarked component
mounted between JK1 and JK2 is R47.
A wire link should be connected between
JK3 (C25 —ve} and JK4 (C26 +ve),
(SOUT...
Please check dates before setting out, as we cannot guarantee the ac-
curacy of the information presented below. Note: some exhibitions may
be trade only. If you are organising any electrical/electronics, radio or
Edinburgh. Al
scientific event, big or small, we shall be glad to include it here. Address
details to Mike Abbott.
$2 Fairs & Exhibitions Ltd., f 01-831 8981
All Electronics Show/ECIF April 30—May 2. Olympia 2. E
Circuit Technology April 30—May 2. E
Field Service & Repairs April 30—-May 2. Olympia 2. E
Automan (manufacturing) May. NEC. Ti
Scotelex June 4-6. Royal Highland Soc., ex. Hall, Ingliston,
Personal Computer World Show Sept. 18-22. Olympia 2.M
International Light Show Jan. 14-28. Olympia. E6 D6 =f 01-701 7127
British Toy & Hobby Fair Jan. 18—Feb. 2. Olympia. D6 E Evan Steadman f 0799 26699
Component Fair March 10. Carleton Community Cntr., Pontefract (on E6 £ 058 84 658
Al to Darrington). F2 F2 ‘Pontefract Am. Rad. Soc. N. Whittingham ¢ 0977 792784
London Medical March 12—15. Earls Court. $2 F5 £ 01-487 4397
IFSSEC (fire/security) April 15-19. Earls Court, London. § I ITF f¢ 021-705 6707
Cast (Cable & Satellite) April 16-18. NEC, Birmingham F5 K2 Reed Exhibitions, Surrey Ho., 1 Throwley Way; Sutton, Surrey
Communications April 23—25. Olympia. I M Montbuild ¢ 01-486 1951
Photoworld April 23—May 6. Earls Court. I i@) Online 01-868 4466
CAD April 26—28. Metropole, Brighton. K2 s f 01-387 5050
Fibre Optics & Lasers April 30—-May 2. Olympia. E Al Institute Electronics ¢ 0706 43661
Custom Electronics & Design Techniques April 30—May 2. E Tl Cahners ¢ 0483 38085
. Practical Electronics: . February 1985
ents in semiconductor technology
introduction of a number of highly
integrated circuits and power Darlington
Ited in audio equipment which is both com-
ty straightforward. This new series deals with the
ion of a variety of modules for use in the custom
gn of sound reinforcing systems and for public address
-work-generally.
©» We start, this month, with full constructional details of a
BOW power amplifier module. This unit forms the basic
building block for several complete designs to be described
later. Details of suitable pre-amoplifiers, line drivers, tone con-
trols and mixers will also be included; the aim being that of
affording the individual constructor the widest possible
choice of audio system configuration.
THE SOW POWER AMPLIFIER MODULE
The power amplifier module is electrically robust, is sim-
ple to construct, and uses low-cost readily available compo-
nents. In its basic form, the module is capable of delivering a
continuous r.m.s, sine wave output of 50W into a 4 ohm
load. The design may be easily modified for operation with
alternative output transistors and/or supply rails, as shown in
Table 1.
Whilst every effort has been made to avoid the pitfalls, it
should be stated at the outset that this project, together with
its higher power derivatives, is not for the faint hearted. In-
deed, the prototype amplifier was not developed without a
few disasters, including four output transistors which literally
melted during the testing stage!
An important requirement of this project (and one which
readers ignore at their peril) is that the loudspeaker systems
employed should be capable of handling the full amplifier
output power. However, readers who do not have immediate
access to correctly rated loudspeakers need not despair
since we shall, next month, be describing a calibrated test
load rated at continuous r.m.s. powers well in excess of
100W. A dummy load of this type should prove to be an
|
|
CAPT Gai?
‘MUTOOLEY ..
D. WHITFIELD
MA MSc CEng MIEE
invaluable accessory for those wishing to ‘run-up’ the
amplifier without destroying their ear drums.
Having started on a cautionary note it is perhaps worth
saying that, provided readers carefully follow the setting-up
procedure and observe the recommendations concerning
heat sinks, component ratings, and supply rails, there should
be few, if any, problems.
CIRCUIT DESCRIPTION
A simplified block schematic of the power amplifier
module is shown in Fig. 1. The corresponding circuit diagram
is shown in Fig. 2, The module consists essentially of a dif-
ferential input stage followed by a driver and complementary
power Darlington output stage. The unit runs from balanced
(i.e. separate positive and negative) supply rails with a com-
mon OV rail at earth potential.
The input stage is formed by TR1 and TR2 which are con-
nected as a long-tailed differential pair with TR3 acting as a
constant current source. The emitter currents of TR1 and
TR2 are determined by VR1 which provides a range of ad-
justment from about 1-5mA to 3-0mA total current. The
signal input is applied to the base of TR1, via a switched d.c.
blocking capacitor arrangement, whilst negative feedback
(both d.c. and a.c.) is applied to the base of TR2. The overall
voltage gain of the module is determined by the amount of
feedback applied and is approximately equal to the ratio of
Max. rec. T1 sec.
heatsink rating
thermal 2x
resistance
TIP126 | 20V/1-54
TIP146 | 20V/2A4
11K80 | 25V/2-54
MJ2501
2N6051
MJ11015] 32V/3A
120W | + 50V
Table 1. Output device selection table
CUSTOM DESIGN
YOUR OWN HIGH
POWER AUDIO SYSTEM
CIFFERENTIAL
INPUT STAGE DRIVER BIAS SUPPLY
NEGATIVE FEEDBACK
(PETSS6P)
BETSETA
SPECIFICATION
Maximum power output: 60W r.m.s. into 3o0hm
{measured at 1kHz) 50Wr.m.s.into 4ohm
40W rm.s. into 8o0hm
25W r.m.s. into 150hm
Minimum recommended load
impedance: 3ohm
Voltage gain: 10
Input voltage for max. rated output: Jess than 2V r.m.s.
Input impedance: 50k approx.
Recommended source impedance: 600ohm
Total harmonic distortion: 0-05% typical at 30W
output into 8ohm
Frequency response (a.c. coupled): 15Hz to 50kHz at—3dB
{d.c. coupled): d.c. to 50kHz at —3dB
less than —85dB
related to max. rated
output
Hum and noise:
Practical Electronics
February 1985
COMPLEMENTARY POWER
DARLINGTON OUTPLT STAGE
Fig. 2, Complete circuit diagram of the Power Amplifier Module
PROIERC :
R3 to R4.
Direct coupling of input signals is
provided by means of S1 which by-
passes the d.c. blocking capacitor, C1.
In order to preserve symmetry of the
differential stage,
resistors are made equal: R2 and R3,
R5 and R6, R1 and R4 (note that the
latter assumes that the amplifier is fed
from a relatively low-impedance
source).
TR4 forms a conventional common
emitter driver stage using an npn
transistor, Since the quiescent power
LOUDSPEAKER
nov
dissipation for this stage is in the region of 125mW, a metal
cased TO5 style device is much to be preferred. Blas for the
output transistors is provided by TR5 which acts as a
constant voltage source, adjustable by means of VR2. The
output stage is a conventional complementary symmetrical
arrangement using Darlington pairs, TR6 and TR7. A variety
of different devices may be employed in the output stage
depending upon output power requirements and the
available supply voltage rails. These configurations are sum-
marised in Table 1, The output stage is protected by means
of two 5A quick-blow fuses, FS2 and FS3. It should perhaps
be mentioned that this form of protection is not completely
foolproof but will normally cope with a short-circuited load
or failure of one of the output Darlingtons,
C6 and R17 form a Zobel network whilst L1 ensures
unconditional stability of the amplifier when operating into a
severely capacitive load. Bootstrap feedback is applied via
C4 in order to raise the effective impedance of the collector
load for TR4. C5 provides high-frequency roll-off since the
bandwidth of the amplifier is otherwise somewhat excessive.
The.power supply arrangement is fairly conventional and
the following,-
provides symmetrical supply rails of nominally +30V and
—30V.
CONSTRUCTION
With the exception of the power supply {T1, FS1, REC1,
C9 and C10) and the output transistors (TR6 and TR7), all
components are mounted on a single-sided p.c.b. measuring
approximatly 65mm x 115mm. The component overlay of
the p.c.b. is shown in Fig. 3. Components should be assem-
bled on the p.c.b. in the following sequence: terminal pins,
resistors, Capacitors, transistors, pre-set resistors, fuse clips,
and inductor, The latter component consists of 20 turns of
20 s.w.g. wire wound with an inside diameter of 8mm. Care
should be taken to carefully remove the enamel at each end
of this component in order to facilitate an effective soldered
connection to the p.c.b.
Sk COMMON
SK1 INPUT
Output power
relative to SW
{dB)
+2b
[PENSSF) Frequency (Hz)
Fig. 5. Frequency response (80hm load)
C9OFVE (+ 30¥)
TRE COLLECTOR
TRE BASE
TR6 EMITTER
SK3 (BLACK)
SKZ{RED)
CIO +VE (O¥)
TR7 EMITTER
TRT BASE
TR? COLLECTOR
C10 ~VE (-30¥)
Fig. 3. Component fayout of the p.c.b.
$K3 (black) SK2(red)
Fig. 4. Wiring diagram for the Power Module
Practical Electronics February 1985
COMPONENTS...
Resistors
4k7 (3 off)
47k (2 off)
470 {2 off}.
10k
220
R10,R14,R15,R16 1k {4 off)
R11,R12 2k7 (2 off}
R13 1k8
R17 100: BW 5% carbon
VR1. 220 min. hor. skeleton pre-set
VR2 1k min. hor. skeleton pre-set
Except where otherwise stated, all fixed resistors are
0.25W 5% carbon.
Capacitors
220n 250V polyester
100 16V p.c. electrolytic
100 63V p.c. electrolytic
220p 25V tubular electrolytic
33p ceramic
100n 250V polyester
C7,C8 100n 100V dise ceramic (2 off)
C9,C10 4700p 63V can elect. {2 off)
Semiconductors bee
D1,02 1N4148 (2 off)
TR1,TR2,TR3 BC212L (3 off}
TR4,TRS BC142 (2 off}
TRE 10K80 (see Table 1}
TR7 11K80 (see Table 1)
REC1 KBPC802 (200V/6A)
Miscellaneous ;
p.c.b. s.p.d.t. miniature p.c. slide switch: :
T1 80VA mains transformer with 220V primary and two
secondary windings each rated at 20V/2A minimum (see
Table 1)
L1 (see text) p.c. mounting fuse clips (4 off)
FS1 24 20mm quick-blow mains fuse and holder
FS2 and FS3 54 20mm quick- blow fuges
Heatsinks (see text)
Silicone impregnated heatsink oannete: (thermal resist-
tance 0-33 deg.C/W) dnd bushes (tyye Sets required)
Terminal pins (13 required)
SK1 5-pin 270 deg. DIN socket: <<: **
SK2 and SK3 4mm sockets (1 red and 1 black}
Mains connector
Printed circuit board (502-01)
The Darlington transistors must be mounted on a sub-
stantial heatsink of no more than 1 deg.CAW thermal
resistance, To facilitate effective heat transfer the use of
silicone impregnated washers is highly recommended (it
should be noted that the collector connections of the
Darlington power transistors are formed by their respective
cases and these will have to be insulated from a heatsink
which will invariably be at earth potential).
The encapsulated bridge rectifier, REC1, also requires
mounting on a heatsink, The requirement for this heatsink is
somewhat less stringent than that needed for the output
transistors and a rating of 5 deg.C/W (or approx. 110mm x
110mm 16 s.w.g. aluminium) should prove to be quite ade-
quate. Happily, with this component, there is no need for an
insulating washer but a liberal applicatiofi of silicone grease
is recommended before assembly. For most practical pur-
Practical Electronics February 1985
All test voltages measured with a 20k ohm/V multimeter.
Table 2. Test voltages
poses the rectifier heatsink can simply be provided by the ex-
ternal case or chassis of the equipment. This expedient will,
however, not normally apply to the output transistors unless
the case is specially designed with heat sinking in mind!
When the p.c.b. wiring is complete, the underside of the
board should be carefully checked for solder bridges and dry
joints, whereas the component side should be examined,
paying particular attention to the correct placement and
orientation of polarised components.
Connections to the heatsink mounted components (TR6,
TR7 and REC1) and reservoir capacitors (C9 and C10)
should be made by short lengths (typically not more than
150mm) of 16/0:2mm (0-5mm/?) stranded pvc covered
wire. A typical wiring layout is shown in Fig. 4.
Internal view of the Power Amp
INITIAL TESTS AND SETTING-UP
Before connecting to the mains supply and switching ‘on’
it is important to observe the following procedure:—
1. Adjust VR1 and VR2 so that they are both in the
fully clockwise position,
2. Switch S1 to d.c. and temporarily short-circuit the
signal input connector, SK1.
3. Connect the loudspeaker (or dummy load described
next month}. The loudspeaker should have an im-
pedance in the range 40hm to 160hm and should be
rated for a continuous power dissipation of 50W.
4. Switch ‘on’ and measure the positive and negative
supply rail voltages. These measurements can be most
conveniently made using the terminal voltages
developed across C9 and C10, respectively. The sup-
ply rail voltages, in the quiescent state, should be in
the range + 27V to + 30V. If the voltages differ ap-
preciably, or if FS1 blows on switching ‘on’, the wiring
OFF
ON
(e
OUTPUT
& POWER Oy:
of the transformer and bridge rectifier should be
carefully checked.
5. Switch-off and disconnect from the mains supply.
Temporarily insert two 100hm 1W resistors in place of
FS2 and FS3. This can be done quite simply by trimm-
ing and folding back the leads of the resistors so that
the body of the resistor is gripped firmly by the fuse
clips whilst electrical connection is achieved without
the need to solder,
6. Transfer the d.c. voltmeter to the output terminals,
SK2 and SK3. Reconnect the mains supply and switch
‘on’. Adjust VR1 for exactly OV. If the adjustment has
no effect or if the resistors get hot, carefully check the
p.c.b. and wiring to the output transistors.
7. Switch ‘off’ and transfer the d.c. voltmeter to the
10o0hm resistor fitted in place of FS2. Switch ‘on’ and
adjust VR2 to produce a reading of 0-2V. Check that a
similar reading is obtained across the 10o0hm resistor
fitted in place of FS3.
8. Switch ‘off and disconnect from the mains supply.
Replace FS2 and FS3 and remove the shorting link
from SK1. Finally, select norma! operation by
switching S1 to the ‘a.c.’ position.
This completes the setting-up procedure and the amplifier
is now ready for use. It is advisable to check the adequacy of
the heat sinking arrangements by observing the temperature
rise of the output transistor after, say, 15 minutes con-
tinuous operation at a reasonable output level (i.e. TOW or
more). If the rise in temperature is more than 25 deg.C
above ambient, the heatsinking should be improved.
NEXT MONTH: We shall provide constructional
details of a 100W dummy load and a simple pre-
amplifier/line driver.
Practical Electronics February 1985
ectrum
DAC /ADC Board |
R.A. PENFOLD
VW {TH something like a million ZX Spectrum computers
now in circulation there are, no doubt, a great many in
the possession of electronics enthusiasts who would like to
use them in computer based measurement and control ap-
plications. One of the ZX Spectrum's main shortcomings is a
lack of built-in interfaces, and there are no ports ready fitted
to the machine that are suitable for applications of this type.
However, it is quite easy to fit interfaces onto the expansion
port, and an analogue interface is one of the most useful
from the electronics enthusiasts’ point of view.
The port featured in this article gives both analogue-to-
digital and digital-to-analogue conversion. Both have 8 bit
resolution, which is more than adequate for most practical
applications, The analogue output has an output voltage
range which is adjustable from O to 2-55 volts to about O to
10 volts, but with additional circuitry the output voltage
range could easily be converted to any desired span within
reason, The analogue input has adjustable sensitivity, with
the full scale value variable from 2-55 volts to about 25
volts. Again, with suitable additional circuitry practically any
input voltage range could be accommodated. The maximum
rate of conversion is guaranteed to be no less than 66000
per second, and in most cases in excess of 100000 per
second can be achieved. Even the guaranteed rate is fast
enough for most high speed applications such as digitising
audio signals.
SYSTEM OPERATION
The block diagram of Fig. 1 helps to explain the overall
way in which the unit functions. The digital-to-analogue con-
verter is the more simple of the two converters. This consists
basically of a precision 2-55 volt reference source, a resistor
network (known as an R-2R network} and eight electronic
switches. The electronic switches are controlled by the eight
DATA
BUS
Fig. 1. Block diagram
Practical Electronics February 1985 ‘
digital inputs, and when activated they connect the precision
reference source through to the output via some or all of the
resistors in the R-2R network. Things are arranged so that
each input, when set high, causes the output to be in-
cremented by the appropriate amount. The operation of this
type of converter has been covered in past issues of this
magazine, and will not be considered in more detail here.
In order to drive the DAC from the data bus of the Spec-
trum an 8 bit latch is needed, so that data written to the con-
verter can be stored in the latch and used to drive the inputs
of the converter: The converter then gives a continuous out-
put, and ignores signals on the data bus that are intended for
other devices. The converter used in this project has a built-
in data latch, and it can therefore be fed direct from the com-
puters data bus. An address decoder circuit provides the
latching pulse when data is written to the converter.
The DAC has.a 2-55 volt reference source, which sets the
maximum output voltage at the same figure. This gives a
nominal O to 2-55 volt output range in 10 millivolt (0-01
volt) steps. A variable gain amplifier enables higher max-
imum output voltages to be obtained, up to a maximum of a
little over 10 volts. Of course, with a higher maximum output
voltage there are still only 256 different output levels, and
the output increments in steps of more than 10 millivolts.
However, for most applications, such as motor speed con-
trollers and even audio applications, the resolution of an 8
bit converter is at least adequate. The amplifier gives the unit
a low output impedance, but without additional buffering
output currents of no more than a few milliamps should be
drawn.
The analogue-to-digital converter is of the successive ap-
proximation type. This incorporates a digital-to-analogue
converter which is driven by a fairly complex control logic
circuit. The eight outputs of this control circuit constitute the
output of the-ADC. The output of the DAC is fed to one input
of a comparator, and the input signal is fed to the other input
of the comparator. When a trigger pulse is received at the
“start conversion” input the most significant bit is set at one,
but the other bits are all set at zero. If the output from the
DAC is at a higher potential than the input signal the most
significant bit is left at one, otherwise it is reset to zero. On
the next clock cycle bit 6 is set to one, and, as before, it is
either left at one or reset to zero depehding on whether or
not the output of the DAC is at a higher voltage than the in-
put signal. On the next clock cycle bit 5 is set to one, and the
process is repeated with this bit. In fact the same process is
used for ail eight bits, and at the end of this procedure the 8
bit binary number fed to the DAC is a valid digital represen-
tation of the input voltage. This method is reasonably fast,
with the conversion taking no more than nine clock cycles,
but successive approximation converters are reasonably
inexpensive,
+12V0
+5V0
z
no R2 VRI
, 390 100k
D190
be 0 16 dl
a; q O OUTPUT
’ 020 1% 1
“| 1c1 22k
D4 O EAE ov
so rt
06 o—
a taal : =e sv
Cia O
ov
RI pa 1C5 = 74.816
-5v0 PIN 14 = +5V
68k 5 VR3 PIN 7 =» OV
bo <—— Ra iM
390 =
w :
4
e 6
oz
a alt . =
ig
2N427
04 @
INF
13
Ds
vile ov
n
07
: as
ca
ov On
me rat
WRO——
ro
0 INPUT
i
Ag O Ic3
74L 5138 his
aed 700k
10REGO— RT
Bie?
¢
ovo
Fig. 2. Circuit diagram of the DAC/ADC board
16
COMPUTING PROJECT Gia
The device used in this project does not have a built-in
clock oscillator, and a simple C-R oscillator is used to
provide the clock signal. The “start conversion” pulse is
provided by the address decoder.. The converter provides its
output via an 8 bit buffer which has three-state outputs, and
it can therefore be connected direct to the Spectrum's data
bus. The ‘‘enabje” pulse for the outputs is obtained from the
address decoder, but an inverter is needed to give a signal of
the right polarity. The converter has a nominal full scale sen-
sitivity of 2-55 volts, but a variable attenuator at the input of
the unit enables this to be reduced somewhat If required.
CIRCUIT DESCRIPTION
The full circuit diagram of the Spectrum Analogue Board
appears in Fig. 2.
All the address decoding is carried out by IC3 which is a
74LS138 3 to 8 line decoder. The Spectrum has a Z80A
microprocessor, but it uses a non-standard method of in-
put/output mapping. The general scheme of things is to have
the address lines normally high, with one of the lower lines
being taken low to activate an input/output device. Some of
the upper address lines are occasionally used to provide ad-
ditional information to an input/output device. This leaves
address lines A5 to A7 free for user add-ons. In this case A5
and the |ORQ lines are fed to the negative enable inputs of
1C3, and A5 must be taken low when reading from or writing
to either section of the port (the [ORQ line automatically
goes low when a BASIC IN or OUT instruction is used).
The three main inputs of IC3 are fed from the read (RD)
and write (WR) lines plus address line A6. This gives four
usable outputs from 1C3, two when reading and two when
writing (four outputs are always high since the read and
write lines never go low simultaneously). This is adequate for
our purposes as only two write outputs and one read output
are needed in this application. When writing data to the DAC
the instruction OUT 65439,X is used, where X is the value
written to the converter. This takes the write and A6 lines
low while the value written is present on the data bus, giving
an output pulse from output 2 (pin 13) of IC3. Other
addresses can in fact be used, but it is best to use 65439 as
this places the address lines apart from A5 and A6 high, so
that unwanted operation of any internal input/output circuits
is avoided.
IC1 is the DAC device, and this is the popular Ferranti
ZN428. It has an integral 2-55 volt reference source, but this
requires discrete load resistor R1 and decoupling capacitor
C1. 1C4 js an ordinary operational amplifier non-inverting
mode circuit, and this amplifies and buffers the output of
IC1. VR1 enables the closed loop voltage gain to be varied
from unity to about 5 or so, but in practice the +12 volt sup-
ply used for IC4 fimits the maximum output potential to
about 10 or 11 volts. VR2 is the offset null control, and this
is adjusted to trim the minimum output voltage of the unit to
zero volts.
The ADC is based on IC2 which is a Ferranti.Z2N427. Like
the ZN428, this has a built-in 2-55 volt reference source
which requires a discrete load resistor and decoupling
capacitor (R3 and C2 respectively). R1 is part of the high
speed comparator, and this is fed from a negative supply so
that comparator will respond properly to voltages right down
to zero volts. R7 biases the input of IC2 to the earth rail and
VR3 pilus R5 are used to provide a small positive bias which
gives improved accuracy at low input voltges. VR4, together
with the input resistance of the circuit, acts as a variable
attenuator.
Practical Electronics February 1985
To EOGE
CONNECTOR |
COMPONENTS ...
Resistors
68k
390 {2 off)
22k
820k
2k2
8k2
100k 0-1W hor. pre-set (2 off)
10k 0-1W hor. pre-set
1M O- 1W hor. pre-set
All fixed resistors are 0-25W 5% carbon
Capacitors
C1,C2 2p2 63V radial elect (2 off}
C3 1nF carbonate
c4 100nF ceramic
Semiconductors
IC1 ZN428E
IC2 ZN427E
IC3 7418138
Ic4 LF351
ics 74LS14
Miscellaneous
Printed circuit board (502-02)
2 x 28 way 0-1 inch pitch edge’connector
8 pin di. ic. socket
14 pin di, ic. socket
Two 16 pin dit ic. sockets
18 pin dil. ic. socket
Ribbon cable, wire, Veropins, solder, etc.
Practical Electronics February 1985
Fig. 3. Component layout of the p.c.b.
IC5 is a 74LS14 hex inverting Schmitt Trigger, but in this
circuit only three sections of IC5 are utilised. One of these
(1C5c) acts as the clock oscillator in conjunction with feed-
back resistor R6 and timing capacitor C3. |C5b merely acts
as a buffer at the output of IC5c. The clock frequency is ap-
proximately 600kHz, which is the maximum guaranteed
clock frequency for the ZN427. However, with most devices
a substantially higher clock frequency is quite acceptable,
and where high operating speed is essential using a
somewhat lower value for C3 to give a higher clock fre-
quency of up to about 1MHz should give satisfactory results.
The “start conversion” pulse is taken from output 6 (pin 9)
of IC3, and is generated using the instruction “OUT
65503,0" (the value written can be any vatid quantity since
the pulse is obtained direct from the address decoder and
not from the data bus). The port is read using the instruction
“IN 65503”. This gives a negative pulse from output 5 (pin
10) of IC3, but this is inverted by IC5a to give the required
positive pulse to iC2.
At least nine clock cycles must be allowed to elapse bet-
ween sending the “start conversion” pulse and reading the
port, to ensure that the circuit has had time to complete the
conversion. There is no problem in BASIC since the slow
speed of this language means that the conversion will
always have been comfortably completed before the port is
read, The situation is different when using machine code,
and it may them be necessary to use a delay loop to prevent
a premature reading of the converter from being taken. The
ZN427 has an “end of conversion” status output, but no
means of reading this have been included in this unit, and as
17
Fe
the length of time taken for a conversion is virtually constant
a'delay loop.is a perfectly practical way of doing things.
- The circuit requires +5, +12, and —5 volt supplies. These
are ‘all provided by the Spectrum from its expansion bus, and
ngormer power source is required.
CONSTRUCTION
“Fhe component layout of the Pree circuit board is
shown in Fig. 3. There are a number of link wires and it is
probably best to fit these first. 22 s.w.g. tinned copper wire
is suitable for the links, None of the integrated circuits are
MOS types, but it is advisable to use sockets for these,
especially in the cases of 1C1 and 1C2 which are not the
cheapest of devices. The integrated circuits do no all have
the same orientation, so be careful to fit them onto the board
the right way round.
Connection to the Spectrum is via a piece of 17 way rib-
bon cable about 0-5 metres long. It is unlikely that 17 way
cable 'will be available, but it is easy to cut down a piece of
20 way cable to the required number of ways. Connection to
the board should not prove to be difficult provided the end of
each lead first has a small amount of insulation removed and
is tinned with a small amount of solder. A 2 by 28 way 0-1
inch edge connector is needed to make, the connections to
the’ expansion bus of the Spectrum. Suitable connectors
complete with a polarising key are now readily available.
Fig, 4 gives connection details for the edge connector.
ADJUSTMENT
Connect the unit to the Spectrum prior to switching on.
The Spectrum should then operate normally — switch off
immediately and recheck all the wiring if it does not.
Assuming ali is well, adjust the DAC first. Set VR1 and
VR2 at a roughly midway setting, and then type the follow-
ing'command into the computer:—
OUT 65439,0
This should give a low output voltage from the unit, and by
adjusting VR2 it should be possible to trim the output poten-
tial to precisely zero volts. Next type into the computer the
command:—
OUT 65439,255
[resva] oc * e5¥ RO? ms “ge” "ps 02. , 00. te
- +12v | WR | ora ya gt. ta
Tri alt
: ” tk
as +5¥
AS ow
Fig. 4. Connection details for the Spectrum edge connector
oe
An output potential of around 7 to 8 volts should then. be
obtained. By adjusting VR1 any desired maximum, output
_ voltage of between 2- 55 volts and about 10 volts or so'can
be set. Repeat this procedure a couple of times to make sure
that everything is set up as accurately as possible. |
To check the ADC arid facilitate its adjustment type in the
following short test program:-— ; .
10 OUT 65503,0 a
20 PRINT IN 65503 = .
30 GOTO 10 , =A
When the program.is run it should return a series of very
low readings (0 or 1). Set VR4 at maximum resistance (fully
counterclockwise), VR3 at a midway setting, and connect an
input voltage to the unit. that is equal to the desired full scale
value. This should be in the range 2-55 to 25 volts. Run the
program and set VR4 just far enough in a clockwise direction:
to give returned values of 255. .
In order to adjust VR3 an input voltage that araetiscill 5
millivolts at pin 6 of 1C2 should be applied to the cirrcuit. In
other words an input potential that is 1/510th of the full
scale input voltage is required. VR3 is then adjusted to give a
series of reading that (more or less) alternate between O and
1. It is not essential to carry out this procedure, and accurate
results will be obtained if VR3 is simply set for about alt
maximum resistance.
SURPLUS to
requirements MC6BO9E,
OSCILLOSCOPE Heath 10-4555 £150' ono
4
SWOP Brothers EP44 computer printer/
typewriter RS232 for oscilloscope. Mr. Small, 8
Cherrytree Road, Chinnor, Oxon.
WANTED Texas Microprocessor TMS 1000.
Please write to: Abbass Rezaei. PO Box 62,
Najatabad, Isfahan, (ran.
FREE-—sacks of old components. Mostly
TV/valves, to be collected in Oldham, Details:
0923 20751, Mr. V: R, Halsall.
FOUR pairs matched boxed speakers, £8 per
pair. T: A. J. Cooling, 4 Norfolk Road, East Ham,
London E6 2NJ.
WANTED Manuals for Cossor oscilloscope type
1035, and Harley oscilloscope type 13A. M. O.
A; Chari, Ladersattravagen 97 3 Tr, 175 70 Jar-
fatla, Sweden.
PAL: information on PCBs making, from
amateur and experts. A. Larry, 56 Becher Street,
Derby DES BNN.
UK101 software for sale or swops. Send for list
of programs. Mr. P. Hale, 31 South Road, Stour-
bridge, West Midlands DY8 3YA.
*"NIGHTRIDER'’ car lights sequencers, drives
nine channels vari-speed. Easy wiring all
negative-ground cars £40 complete. Mr. Ss. M.
Budzinski, 16 Laburnum House, Malpas Road,
London.SE4 1BL.
SN74LS783 Synchronous address multiplexer
chips. £4.00 each. 10+ £35. SN74LS783 data
£2.50. Mr. N. E. Spiers, 114 Green Way, Tun-
bridge Wells, Kent TN2 3JN.
CLEARTONE graphic equaliser (battery) 7
channel mono, with master volume.control, New
£22. inc. p&p. F, C. Smith, 283 Leeds Road,
Newton Hill, Wakefield WF1 2JQ. Tel: 0924
374122.
40 x B255A £70 the lot or £8 each. Also printer
leads for Dragon 32, BBC £9.95 each.
R, Vowles, 3 Orchard Waye, Uxbridge, Middx.
UB8 2BN. Tel: 0895 54720.
WANTED two SN76001 i/c as used in the
Heathfield TV. Mr. Kendall, 4 Howlets Terrace,
Chelmondiston, Ipswich IP9 10X.
5 x 7 dot Matrix printer 7 colours: 80 cols.
parallel interface adjustable tractor feed VGC
£220 o.n.0. S. Walker. Tel: O865 750600 even-
ings.
CAR battery voltage monitor — graces any car.
As new in immaculate condition £3.45. Russell
Oakes, 32 Wigan Road, Winstanley, Wigan,
Lancs. WN5 7XS.
COPIES available from private collection early
service sheets radios TVs etc. £1 + large SAE.
State make, model. Maurice Small, 8 Cherry Tree
Road, Chinnor, Oxon, OX9 4QY.
HAMEG oscilloscope HM203-4 dual beam
20MHz with probes £195. Tel: Southampton
557386. Mr. D. Couchman,,8 Grosvenor Gar-
dens, Southampton.
- Richards,
also other instruments P.C. bridge etc. Offers,
good condition, sold separately. Mr. A. Ewing, 9
Croft Crescent, Markinch, Glenrothes, Fife KY7
6EH, Scotland. | 2 rps
WANTED AY-3-1270 linear i.c. oor
RS-3-1270. Mr.. J. F. Wilson, 233 Broomlee
Close, Newton Aycliffe, Co, Durham. Tel: Avett
312130.
WANTED two track record head for B & ¢.
record 1800 RTOR deck, Mr. C. Bressington, 47
Station Road, Ystrad Mynach, Mid- -Glamorgan.
Tel: 0443 813005.
EIGHT Philips LVC 150 25hr. video tapes. Har-
diy used, Offers plus postage. Mr. L. T. Hill, 14
Rothesay Terrace, Bedlington, Northumberland.
Tel: Bedlington 825967.
WANTED data or specimens of early transistor
types. Good prices paid. Write for full details: Mr.
Andrew Wylie, 18 Rue de Lausanne, 1201-
Geneva, Switzerland.
MICROSYNTH Synthesiser built and tested
with speaker and homemade stand £150, 8
Stourton Road, Witham, Essex, Tel; Witham
514556.
COMPONENTS transistors mainly OC/AC/BC,
capacitors, resistors, chips, pots, relays, motors
and other, 16 |b, £20. Mr. Turner, 4 Mill Fields,
Newtown, Powys. Tel: 0686 27862.
WANTED service circuit diagrams Sugden C51.
A51 purchase hire to copy. Good price paid.
Maesyffynnon, Caehopkin Road,
Abercrave, Swansea. Tel: 0639 730629.
Practical Electronics February 1985
oy
Outlook
Everyone in business breathed a sigh of
relief once the US presidential election was
over. The two-year run-up is almost un-
bearable for its unsettling effect.
All indications are that 1985 will be a
good, though possibly hard year for the
electronics industry. In 1983 the number of
small companies starting up provided a net
gain of 47,000. The 1984 figures, not yet
available, are expected to beat this record
and there is no reason why 1985 should be
worse. A significant number of the new
start-ups will be in or associated with the
electronics industry.
Foreign investment in the UK continues
at a high level. The prolonged miners’ strike
was apparently seen overseas as a one-off
out-of-character industrial relations
problem and not typical of the ‘new realism’
in British industry. In any case no foreign
companies would want to invest in coal
mining and potential investors may be im-
pressed by the relative ease with which
British industry carried on through what
was intended to bea crippling exercise.
Expansions and new starts planned last
year will begin to take effect. If we look at
Scotland's Silicon Glen there are now
nearly 300 electronics companies employ-
ing more than 40,000 people. Inward in-
vestment since the Locate in Scotland
agency was founded in 1981 has now top-
ped £1,000 million, most of it finding its
way into high-tech projects. And substan-
tial investment is similarly going into other
areas. Even so they will not create many
new jobs, one recent estimate being for
10,000 in the electronics industry this year.
Intense competition in personal com-
puters will force prices down so intending
buyers would probably profit by delaying
purchase. Alternatively, prices could
stabilise but the product improved in the
classic ‘‘more-bits-per-buck’’ context. But
despite all the difficulties it will still be
possible to score well in the consumer
market as proved by Alan Sugar’s Amstrad
whose pre-tax profits topped £9 million in
its last financial year.
Practical Electronics February 1985
Information Technology
Information Technology which ‘seemed
so novel {although hardly new} only three
years ago has now become accepted as the
norm and no more exciting than radio or
television. Some pessimists are already
saying that Britain has been losing ground
in this growing sector of industry.’ Their
fears are based on the increase in imported
equipment compared with indigenous
production.
At the higher jevels of the technique the
Alvey programme is now gaining momen-
tum. Two new major contracts were awar-
ded towards the end of last year. One
carries the painfully contrived acronym of
ADMIRAL derived from ADvanced Mega
Internet Research for AtLvey. The other is
merely called the Speech Recognition Pro-
ject.
The ADMIRAL contract is a £3 million
joint venture. being coordinated by GEC
Research Laboratories! Partners are Univer-
sity College, London, The University of Lon-
don Computer Centre and British Telecom
Research Laboratories.
The system will link local area networks
(LANS) through a ‘mega internet overlay’ to
produce a single large system of data
networks. The key feature is to allow high
speed intercommunication between dis-
similar equipment. It appears to be a re-jig
of part of the Project Universe programme
originally initiated by the Department of
Trade and Industry and later transferred to
the Alvey Directorate.
The voice recognition project is funded
with £2 million and is centred on British
Telecom Research Laboratories. with
collaboration from Cambridge University
and Logica. Although most schoolchiidren
are acquiring keyboard skills it is also
recognised that speech is the most natural
form of communication between people
and the same applies between people and
computers. .
Present voice recognition systems are
primitive and generally respond = only. to
single-word voice commands. It is hoped to
expand into true verbal dialogue between
person and computer so that anyone who
can talk can, for example, find what is re-
quired from a data base without necessarily
having any keyboard skill.
Time for Schools
It is heartening to see that schools are
-now to be networked through The Times
Network for Schools (TTNS) which will give
any school access to more than 200,000
pages of information by the end of the year.
Secondary schools of which there are
6,500 will be first to join, followed by.
27,000 primary schools.
Future plans exist to network British
schools to those on the continental!
mainiand. Joining the network will be op-
tional but the fees are modest and the
scheme should prove popular and exciting.
Then we have the proposal for a univer-
sity devoted entirely to information
technology. It has the backing of a host of
leading electronics companies but students
will have to pay fees to make the university
self-financing. °
Fears have already been expressed by,
egalitarians that the university will be anti-
social because it will create an elite of the*
already advantaged who can afford the
fees. Nonsense, of course. We need more,
not less, centres of educational excellence.
Improvisation .
| remember at a press conference:
organised by a Ministry of Defence elec-
tronics establishment asking how we would
get on in a real war when the time scale of
equipment development stretched over a
period of years even when the equipment.
was comparatively simple. | had in mind the
very few weeks which elapsed in 1940 bet-'
ween the alarming discovery that the Ger-
mans were using magnetic mines and the
countermeasure (degaussing of ships)
devised and implemented. And this was.
only one example of many rapid develop-
ments of that war. ;
{ contrasted this with over 10 years of
development of the Clansman radio system.
before it came into service. | was not very
impressed with the reply which was more
or less that we would probably muddle-
through as we always had done in the past.
My confidence has now been restored.
(well, almost) by Alfred Price in his new
book ‘Harrier & Sea Harrier at War’,
published by lan Allan Ltd. In it he
describes ‘Blue Eric’, an electronic counter-
measure system needed urgently for the.
Falklands war. If Harriers were to’ bé
operationally successful in the South Atlan-
tic they would need self-protection against
Argentine radar installations.
The threat was evaluated from normal
military intelligence which already had
details of the characteristics of radar equip-
ment in service with Argentine forces. Ex-
isting electronic warfare pods (e.g. for Tor-
nado and Buccaneer) were too large and
heavy for the Harrier so it was decided to
use elements of the Sky Shadow equip-
ment and fit them in a modified gun pod
which would meet the weight and size re-
quirements as well as the jamming
capability.
Marconi Defence Systems were prime
contractors and completed the design,
testing and delivery of operational units
within 15 days instead of an estimated two
years at normal pace and at a quarter of the
cost.
Blue Eric (named after its MOD project
officer Squadron Leader Eric Anna!) was
never used in the Falklands. When the EW-
equipped Harriers arrived they were groun-
ded for four days by bad weather and by the
time they got airborne the conflict was vir-
tually over, the Harriers then being used for
front line ground attack where the radar
threat was negligible or non-existent. t
While it is comforting to know that the
improvisational skills of yesterday have not
been lost, one is still left wondering why an
EW pod for Harrier was not already
available and why, in peacetime conditions,
equipment development times are so long
and the cost so great.
21
The Principle and the Product
HE electronics industry is one of extremely rapid change.
New products, ideas, and standards spring up continually,
promoting a continuous state of flux and development. As a
fairly young industry it is successful in discarding old and out-
dated principles in favour of newer, more beneficial ones; if
change can be shown to be worthwhile in any specific situation,
then that change is almost invariably made.
It comes as somewhat of a surprise, therefore, that for the
production of documentation, text, correspondence, and com-
puter programs, the primary means of interface between the
human being and the machine is still a QWERTY keyboard.
QWERTY is the standard layout of typewriter keys which was
devised very many years ago with the principle intention of
Slowing ‘Down the typist to prevent jamming of the mechanism.
In this age of mechanical sophistication and electronic
keyboards the same requirement is no longer true since we can
easily prevent jamming by other means. Hence, we are left with
a legacy from a bygone age. The QWERTY layout is slow and
complex to learn, with months of training being required before
any proficiency is achieved. For many people such training is
impractical, so they are reduced to ‘two finger’ typing, which is
usually a slow and frustrating exercise.
A NEW IDEA
When a company brings-forth a new idea for entering text
into machines, it is bound to attract considerable interest. A few
years ago a device called the ‘Microwriter’ appeared. It is a
small, self-contained machine with only six keys, which is used
with one hand only. The Microwriter company has been produc-
ing these devices in modest quantities ever since, and has
recently started to advertise and promote the product in a more
aggressive way, with various options and accessories now
available,
WHAT IS A MICROWRITER?
A Microwriter is no less than a battery powered portable
word processor. It is just a little larger than a paperback book
and has very few controls—some connectors, an ON/OFF
switch, a liquid crystal display, and six keys. It is placed on a
desk or held in the left hand, and typed on with the right hand.
(As yet there isn’t a left handed version since there would be
problems connected with the way that the ‘alphabet’ of letter
shapes are formed, as we shall see later.) The Microwriter can
communicate over a bi-directional RS232 serial link with
printers, full-sized word processors, computers, etc., and can
store text either internally on battery backed-up RAM, or on
any conventional external cassette recorder.
Characters or numerals are entered into the machine by
pressing combinations of keys, rather than one key at a time as
in the case of the QWERTY system. There are no markings on
the keys since they can have different functions at different
times, so the user is immediately forced into the excellent prin-,
ciple of touch typing, and looks at the display rather than at the
keys being pressed. The user, therefore, has to learn all the
sequences of keys to be pressed before being able to type cor-
rectly. This is the make-or-break aspect of the Microwriter—
many people are immediately put off by having to learn a poten-
tially complex typing language. Fortunately, the people at
Microwriter Ltd. have been very clever indeed in the choice of
22
Tom Gaskell sa (Hons) cEng MIEE
keys to be pressed per character. The right hand is always held
in the same place above the keyboard, one finger above each
key, and the shape formed by the fingers pressing the keys bears
a relationship with the shape, or some aspect, of the character -
which becomes entered into the machine. That relationship is
sometimes obvious and direct, sometimes humorous, sometimes
very corny, but inevitably is easily memorable. Fig. | shows
some of these relationships, based on a slightly stylised layout of
keys, The manufacturers suggest that they can be memorised in
typically one hour, and certainly I have found this to be the case
as far as friends, colleagues, and myself have been concerned. It
is very easy indeed to learn to Microwrite; far, far easier than
touch typing, and I have tried both!
USING THE MACHINE
As each character is entered from the keyboard it is displayed
on a single line liquid crystal display which can show up to 14
characters and two control symbols. The display acts as a
‘window’ on the text, and can be moved around within the text,
either following new characters entered or to review what has
already been written under. the control of special commands,
The text, as shown by the display, normally appears to shift to
the left as each new character is entered by the keyboard and ap- ‘
pears at the right hand end of the display. The sixth key on the
Microwriter is a second thumb key, a little below the normal
one, and it acts as a control key, allowing comprehensive control
of the machine’s functions. It is used either on its own, or with
other keys in place of the normal thumb key. For example, en-
tering the letter ‘f (the First Four keys pressed), but using the
control key instead of the thumb key, moves the display window
in the text Forward one position; ‘f for Forward—it is corny,
but it works! Doing the same thing with the letter ‘k’ moves the
window backwards. In this case, ‘k’ stands for Korrect, so you
use it when Korrecting errors!
WILL THIS
EVER REPLACE
THE ‘QWERTY’
KEYBOARD?
i
.
Practical Electronics February !985
Pressing the control key once, on its own, puts the
Microwriter into upper case characters for just one entry, after
which it reverts to lower case. Pressing it twice in succession, on
the other hand, locks the machine into upper case continuously
until the two thumb keys are pressed together to revert to lower
case. The status of the machine is continuously shown by two
control symbols in a yellow coloured area at the right of the
display.
Simple punctuation is provided as part of the normal lower
case letter set, but more complex punctuation and numerals
have to be accessed by a ‘numerals shift’ function. Entering the
letter ‘n’, but with the control key pressed at the same time, shifts
the Microwriter into the numerals mode for one character only;
entering this combination twice in succession locks in the
numerals mode, just like the upper case mode. There’s another
set of character/key relationships for the punctuation, with num-
bers being entered by a ‘count on the fingers of one hand’ type of
technique. The requirement to shift for numerals is acceptable
for word processing applications, but would make the
Microwriter somewhat laborious for writing computer
programs, for example.
To avoid timing problems when keys are pressed together, the
Microwriter works on a key accumulation principle, and the
character is only entered when all the keys have been released.
Hence, you can start to press keys in any order, so long as at
least one of them is being held down at any given time. When all
the keys are eventually released, the result is as if all the keys
that were depressed in that sequence, irrespective of their
chronological order, were pressed simultaneously. This makes
the keyboard action very ‘forgiving’, and allows characters to be
entered very slowly and deliberately when required. The speed
which can be obtained after only a few weeks’ use is very high—
not as fast as touch typing, but certainly up to twice as fast as
handwriting.
Practical Electronics February 1985
EDITING AND WORD PROCESSING
When text has been written it can be edited (both deletion and
insertion) and reviewed by appropriate use of the control key.
To read through the written material, the user has to Jump back
to the beginning (control + j) then scroll Forwards (control on
its own, followed by control + f); this then moves the display
window along the text one word at a time, at a user selectable
slow or fast rate, until you tell it to stop. The machine
automatically enters ‘carriage returns’ at the end of each line,
and ensures that these are between words, not in the middle of
them. Via the control key the user can access tabulation, margin
indents, document markers, page separators, alter line length,
and do many other complex word processor functions. These
become very difficult to memorise, and even more difficult to im-
plement, and I would have thought that they would have only
limited usefulness to most people.
The control key is also used to suitably configure the RS232
link. Although this can be used to Joad text into the Microwriter,
its primary use is to transmit text to a computer, word
processor, or printer from the Microwriter. Full handshaking is
provided, and there are user selectable baud rates, data lengths,
etc., so it will interface with most RS232 based systems. All
settings and text are stored in RAM with battery back-up, so
nothing is lost when the power is turned off, The machine even
turns the power off itself if it is not used for a few minutes, to
CUO necoenbep
e
microwriterl LEARNING
] SEQUENCE
Fj Be } Ps,
°
rome
| H| @)
Top of tneT
Fest Four Most Fingers
Finga‘sforF FM Radio Make
completely for P
Go 0°
n a
- FAC )
O fp My Le
a C
D [oea SH / es
finger lindex| for The centra: target Signet ting } ais
Space most common letterE| - bulls eve * finger Very non-U
ip ob -
Adyotning
The dome of theO The bump of theB First upstroke downsiroke
Looks liken ¥ of the of then
Either side at the common line
——
The two sides
of the Ww
First downstroke
ofthe
The upstroke
ofthe
Make a tai from
the central
a wee | 5(/ 5 o ( —¥
|
Everything Zig zag between Full stop. :
Xcept your index the keys forZ come to @ point Hyphen Comma
Command Key
y | 0%
re) Oo
©
Apostrophe
Pairs of tetters have Deen
highlighted by outlining
sr en if 1 = ae
84228-1 (11/83) © Copyright Microwriter Ltd 1982 07-022
Fig. 1. Sometimes corny, but inevitably memorable
23
conserve battery life. The batteries are rechargable types, and a
suitable charger is provided with the machine. Up to 1600:
words, or typically 5 pages of A4 size, can be stored in. es
memory of the machine. (Much more if cassettes ‘are used.) ..
THE HARDWARE
The packaging of the Microwriter inspires confidence! It is
housed in a very solid injection moulded plastic case. The keys
are ultra-low activating-force microswitches with moulded keys.
Their action is light but positive, and their positioning is
ergonomically spot-on. Inside there is just one main p.c.b.
holding the RCA CDP1802A CMOS microprocessor, four
HM6116 CMOS 2k Byte static RAMs, and a 2564 8k Byte
CMOS EPROM, along with an ‘intelligent’ liquid crystal display
above it as a sub-assembly, and other assorted CMOS i.c.s. The
batteries are housed between the microswitches in the upper half
‘of the case. It’s a well laid out and professionally built product.
With the Microwriter itself comes a good quality soft carrying
case, a battery charger, a cassette recorder connecting lead,
some ‘crib cards’ giving a quick reference to control codes,
characters, punctuation, etc., and two instruction manuals; a
new user’s guide, and a more complex systems manual for
setting up communications protocol and the like. The new user’s
guide is effectively the main instruction manual for the machine,
and without doubt is the best manual that I have seen for a
piece of consumer electronics. The cartoon characters used
might annoy some, but they will drive the points firmly home to
just about anybody, whatever age or ability. Other product
manufacturers would do well to study this manual and compare
its high standards with their own!
There is an optional television interface unit available for the
Microwriter which I’m somewhat less happy with. It interfaces
to the RS232 port, and allows the display of text on a domestic’
television set or a composite video monitor. It is expensive
(around the £100 mark) and gives very limited facilities. Writing
onto the screen as you enter text works reasonably well, but if
you just want to dump a letter, for example, onto the screen to
check its layout, the use of the Microwriter becomes somewhat
more contrived. It’s very difficult to put a letter onto the screen
without the top of the letter scrolling off the screen as soon as
the bottom of the screen is reached. The unit that I tested also
failed to get the ends of the lines correct when dumping onto the
screen; parts of words were left at the end of some lines, then the
whole word reappeared again on the next line. For the majority
of potential Microwriter users I would question the necessity for
the television interface unit—when you've got the hang of
Microwriting you probably don’t need it. It seems to be more
suitable as a shared facility between several users, and generally
seems to be somewhat of an afterthought rather than an integral
part of the Microwriter system.
THE QUINKEY
For many people, the cost of a Microwriter (£299 plus VAT),
although low by office equipment standards, is too high for them
to consider it as a personal purchase. However, they could con-
sider investing in a ‘Quinkey’. This appears to be an ordinary
Microwriter at first glance, but lacks most of the connectors and
the display. In fact, it contains no electronics, just a set of
microswitches and resistors which enables up to four of them,
ingeniously, to plug into the analogue inputs of a BBC
microcomputer. For just under £50 the full Quinkey package
provides good value, consisting of the the Quinkey itself, a
manual, some crib cards, a connecting lead, and the software to
run the system. Further Quinkeys on their own cost around £30.
The software enables the Quinkey to be used as well as the
standard BBC QWERTY keyboard, not only with software
within the BBC micro such as BASIC, the Acorn DFS, etc., but
24
also with software packages such as Wordwise and similar. A
“version for the’ Spectrum is .soon’ to be* made available, and
_ -Microwriter are: working on yersions for other popular personal
computers too. All this helps * to bring the, unique. qualities of
Microwriting to the private individual, schools, colleges, etc.
APPLICATIONS—WHO USES IT?
The most obvious market for ‘the Microwriter is with
professionals on the move—salesmen, executives, engineers, and
anybody who does an amount of documentation, report writing,
letter writing, etc. On the train or ‘plane they can write their
meeting reports, or they can keep notes in the field or by their
work benches, and either print the results out so that they are
legible to themselves and to their colleagues, or if necessary
dump them onto the office computer ‘or word processor to be
tidied'up before final printing. There’s no duplication of effort,
the typists no longer having to work from handwritten notes.
The small size, portability, and easé of use of the Microwriter
are attractions which a QWERTY keyboard has never. had,
Microwriting can never be as fast.as good touch typing, so it will
not be used to replace QWERTY keyboards in typing. pools or:
‘secretarial offices, but for thousands of unqualified typists. it of-
fers a refreshing alternative to the two-fingered struggle, so it
should be of great interest to small businesses, the police, sales
personnel (especially those working from home), budding
novelists, and even to the writers of magazine articles! For
schools it has the advantage of allowing the connection of four
Quinkeys to each BBC microcomputer, which immediately
shares out normally limited resources to many more children. If
accepted for these applications, it can only help. establish
Microwriting as a world-wide standard in years to come.
THE FUTURE OF MICROWRITING
Until recent months the promotion of the Microwriter was a
very low-key process, although some rather more prominent
advertising is now being seen. Over 7000 have been sold, which
can only be the very tip of the potential iceberg. I must express
reservations, however, about the approach that Microwriter are
making on the market place, which seems to be rather uncertain
and lacking in self-confidence. I first saw a Microwriter ‘in the
flesh’ in the latter. part of 1983, when I had a demonstration and
a loan from a distributor for a couple of weeks. I expressed a
great interest when I returned the machine to him, and was
promised more information and a follow-up call shortly. I never
heard from him, or another distributor I contacted, ever again.
At the end of January 1984 I approached Microwriter them-
selves for information and a review sample to help prepare this
article. I also ordered two Quinkeys for my own use. I am
writing this article in mid August; the review sample only arrived
three weeks ago! The Quinkeys arrived in the middle of June,
some 19 weeks after they were ordered, and only after telephone
calis at the rate of once per fortnight for most of that period. I
persevered—I wonder how many others did not?
I hope that the future is very rosy for Microwriting. Amongst
friends and colleagues the Microwriter has created more interest
than any other piece of equipment that I can remember. The
concept of the Microwriter is a work of genius. The market is
potentially vast, the product works well, and the presentation is
superb. The price is a little high, but should not deter the
professional market, with the lower cost market being satisfied
by the Quinkey. Let's just hope that Microwriter can improve on
the delivery and planning side of it, put some more aggression
into the marketing, and produce a commercial, not just a
technical winner, What a great shame it would be if the
Microwriter concept was lost to an overseas supplier, as has
happened to so many other viable products from the UK.
More information can be obtained from Microwriter Ltd., 31
Southampton Row, London WC 1B 5HJ. (01-831 6801).
Practical Electronics . February 1985
THE LEADING EDG
DIGITISATION
Everyone talks about the information ex-
plosion, The key is digitisation. With digital
telephone systems what goes down the line
is a series of PCM pulses, rather than
analogue waves. Once you have that situa-
tion, the sky's the limit.
PCM pulses can carry telephone quality
speech, high quality stereo radio, TV pic-
tures, computer data, teletext, viewdata; in
fact any information that can be converted
into an electrical signal. Switching is by
microchip, instead of the primitive
Strowger electro-mechanical relay which
phone systems have used for the best part
of a hundred years.
By interleaving different calls in the same
data stream, the capacity of a link goes up
around 15 times, i.e. a pair of copper wires
that normally carry one analogue telephone
call, can carry fifteen digitals. With optical
fibres, and the signals carried as light
pulses rather than electrons, capacity rises
much, much higher.
The British Post Office started working
on PCM phone links 20 years ago. Few
people know that the PO installed an ex-
perimental digital exchange at Earl's Court
in 1968 and had it running until 1975. That
was when talk about System X started.
Cynics say that the System was called X
because no-one really knew what it was
going to do or how it was going to do it.
Essentially it's a computer switching ser-
vice for PCM streams and there are now six
System X exchanges working in London.
One is at Baynard House in the City of
London, The first five were prototypes.
Once data streams are digitally switched
and connected, the options available open
up wide. There is no problem in providing
conference calls, automatically re-directing
calls to other numbers or displaying the
telephone number of origin when you
receive a call.
ELECTRONIC MAIL
Already many people in Britain are using
electronic mail, which is a hybrid system of
sending digital data down an analogue
telephone line. I'm one of them and there
are quite a few stories to tell about how the
system works in practice, as opposed to
theory!
More of that in a future month. At the
moment | am trying to find out why the
main computer used by Telecom Gold for
electronic mail keeps going wrong and
leaving users like me stranded!
Why worry about information
technology? There's a very short answer. /t
is always far cheaper to send electronic
data down a telephone fine, or over a
wireless link, than shift people or bits of
paper from town to town or country to
country.
Practical Electronics February 1985
The best example of this is what happens
at the Economist magazine. This London-
based publication also prints in America
and the Far East. Printing master plates are
sent by airline courier to the Orient. Until a
year ago they were also sent to America.
The plum job on the Economist was to
take a day trip on Concorde to New York
and back, with the print plates, for safe
keeping. Now the magazine text is con-
verted to digital data and sent by satellite
direct to a Connecticut printing works,
which publishes virtually simultaneously
with London.
Wisely the Economist still sends a back
up text by ‘plane just in case the satellite
link breaks down. But no-one gets the plum
job of going along with them any more.
VIDEO NEWS
Polaroid has joined Kodak in 8mm video.
Sony may follow next year but so far
everyone else is sticking with their existing
VHS and Beta formats. Ironically by joining
Kodak, Polaroid may well have helped its
rival succeed. The extra name gives the
new format credibility.
‘At the Chicago Consumer Electronics
Show both companies were demonstrating
NTSC camcorders using the 8mm cassette.
Picture quality was good and sound, using
f.m. mono, seemed OK. The big question
mark is over tape supply.
Video writing speed is very low; 3.8
metres a second for NTSC and 3.1 metres a
second for PAL and SECAM. So packing
density must be very high. You can get it
either from tape coated with metal powder
{MP) and coercivity around 1600 oersted.
But this needs video heads which are ex-.
pensive and may be short lived. The other
way is to use lower coercivity tape coated
by evaporation of cobalt-ferric metal in a
vacuum (ME). No one has yet succeeded in
making ME tape reliably in bulk.
Kodak started shipping 8mm camcorders
to US traders last September. A 90 minute
cassette costs $24 and the system $2000.
There is no sign yet of a PAL or SECAM
prototype. Although 8mm video almost cer-
tainly comes too late and too expensive to
catch the domestic market, it could well
form the basis of a new professional cam-
corder format.
Sony has both domestic and pro in-
terests. Kodak and Polaroid are paying
Matsushita, Toshiba and TDK to get the
technology right for domestic use;
Professional use is the logical follow on. -
CLEAN CUT
| have now seen inside several compact.
disc and videodisc manufacturing plants.in
Britain, Germany and Japan. They all -have
one thing in common with a microchip.
factory, that is absolute cleanliness.
_parable steps to preserve cleanliness. The
@'second, ‘ to “give a® constant tracking,
“are 4-5 kilometres of track on a single si
Exactly the same situation exists: in
magnetic tape factories, where any dirt,in
the atmosphere will end up as non-
magnetic blemishes in the coating and
cause dropout,
Air in the so-called “clean areas” is
filtered to Class 100, that is to say less thary:
100 particles of less than 0-5 micron size in
every cubic foot of air. The pressure-of air
inside these clean areas is higher than the;
atmosphere outside, so when a door opens
clean air blows out and dirty air leaks in, :
The staff must wear full length, lint-free.
jump suits, like space clothing, and only a
few visitors are allowed in. Usually there is*
an air shower, where blasts of clean air
flush dirt, dust, dead skin and dandruff off
every human passing through.
If only, | think every time | visit-one a
these plants, factories which press ordinary
records would take even remotely com-,.
official answer is that it's not necessary.
Certainly, by comparison, the technology
of LP production looks like a blunt instru- -
ment. But it is easy to forget that a viny! LP’
record is by far the most precise product
mass produced from plastics!
The groove of an LP record is specified’
by IEC standard to be never less than 25::
microns (or millionths of a metre) wide.and.-
preferably not less than 35 microns wid
As a “yardstick” a human hair is around 50
microns in width, The IEC puts stylus tip.
radius at between 15 and 18 microns.
Now let's look at a Laservision videodisc, *
and a compact disc digital audio record,
Both have a spiral of information pits with a
track pitch of 1-6 microns. : pooh gS
For videodisc the pits. are 0-5 microns!
wide, and for compact disc they. are: 0+
microns wide. Video pit depth. is 0.1 micron ie
and CD depth 0-12 microns. :
In other words there is very little dif:
ference in the dimensions; both are at least’:
50 times smaller than the LP groove. The:
laser spot for videodisc ‘playback ‘i
focused to a circle of 0-9 micron diameter:
and for compact disc itis ‘1 micron,’ The:
layer of protective lacquer in a compact.
disc has to be exactly 1-2 millimetres thiek)””
or it will affect the laser focus,
PARTY TURN
- Ifyou collect useless information to bring
the’ ‘disc ata speed which varies between
3:5 revolutions a second and 8 ‘revolutions
velocity of '1-25 metres a second. i”
: That means that for a one hour dise there
i a laser videodisc ‘the track eet iy
SE
Cl
POWER OP-AMPS (TCA365 and TCA2365)
E of the most important components
available to the analogue circuit designer
is the operational amplifier, or ‘op-amp’, The
majority of these, however, are somewhat
limited in their load driving capabilities.
Simple devices such as the 741 can only out-
put 25mA under short circuit conditions, or
10mA in normal operation. For much higher
currents it is usually necessary to add extra
driving transistors to a conventional op-amp.
The TCA365 and TCA2365 are power op-
amps which allow the designer to use a single
ic, in high power applications rather than the
more cumbersome ‘op-amp plus-components’
approach, In practice, they behave as fairly
ordinary op-amps with the exception that their
output stages can drive up to 3 amps in the
case of the TCA365, or 2-5 amps per
amplifier in the case of the dual op-amp
TCA2365. The two i.c.s are very similar, with
both sets of specifications being given in
Fig. 2. The main points to watch are supply
voltage maxima, output currents, and power
dissipation, these all vary between the 365 and
the 2365, (Note that the output current shown
for the TCA2365 is 2:5A per amplifier, not
for the whole i.c.) Fig. 1 shows the pinouts of
the i.c.s. For moderate to high power applica-
tions, heatsinks should be used. These should
be insulated from the i.c.s’ tabs if the internal
connections to the —ve supply could cause
short circuits or problems.
The TCA2365 has an ‘inhibit’ input which
can be used to turn the outputs of the op-amps
off, ie. high impedance (approximately 4k). In-
hibiting is effective when pin 6 is taken to the
—ve supply rail, and the amplifiers operate
normally when pin 6 is taken above 3-0V
referred to the —ve supply, or left unconnec-
ted. Both the TCA365 and the TCA2365 have
extensive protection; they are d.c. short circuit
proof and have thermal overload and safe
operating area protection. The internal current
limiting makes them ideal for driving complex
loads, and especially for driving filament
lamps, whose low resistance in the ‘cold’ state
can cause problems with other types of driver.
BASIC CIRCUITS
Some basic circuits for use with these
power op-amps is shown in Fig. 3. In all cases
there is an external Zobel network (sometimes
known as a Boucherot network) fitted between
the output and O volts to help to maintain
stability under widely varying load conditions.
The | ohm resistor does not have to be high
power (4 watt will do) and the capacitor must
be 100nF for the TCA365, or 220nF for the
TCA2365, It is unimportant which way up the
network is fitted; the capacitor can be connec-
ted to OV and the resistor to the output, or vice
versa. Both power op-amps can be used with
either split or single rail supplies, just as would
be possible with most conventional op-amps.
Figs. 3a and 3c are very straightforward
conventional op-amp circuits, and apply per-
fectly well to the TCA365 and TCA2365. For
minimum offsets, Rg in Fig. 3a and R, in Fig.
3c should be included as shown, although in
many circuits these are unnecessary and can
be replaced by short circuits for economy.
Both these circuits, however, should really
only be used for higher gain circuits; + 10dB
or more, or preferably +20dB. For lower gain
circuits, and certainly for anything less than
10dB (approximately x3), the configurations
of Figs. 3b and 3d should be used. For unity
gain, use typically between 10k and 100k for
both Rj and Ry, with Ro approximately one
tenth of that value, in Fig. 3b, and typically
between 10k and 100k for Re in Fig, 3d, with
Ro one tenth of that and Ry an open circuit.
The reason for all this is concerned with
stability.
INPUT
-VE SUPPLY Cott
OUTRUT rr
+VE SUPPLY Creer
OUTPUT
INVERTING INPUT
NON- INVERTING INPUT
-VE SUPPLY
+VE SUPPLY
INHIBIT INPUT
NON- INVERTING INPUT
INVERTING INPUT
OUTPUT
OP-AMP a
OP -AMP b
NON-INVERTING
STABILITY
There are many factors influencing stability
in operational amplifiers. These tend to be in-
volved, complicated, steeped in complex-plane
mathematics, and certainly beyond the scope
of Semiconductor Circuits! Empirically,
most electronics engineers and enthusiasts
learn some straightforward rules of thumb
about how to keep amplifiers stable and
prevent problems of self-oscillation at
several megahertz. A common ‘cure-all’ is to
connect a small value capacitor, typically less
than 100pF, between the output and the in-
verting (—ve) input. Don’t do this to a
TCA365 or 2365! Even if it doesn’t actually
cause oscillation (which it probably will) it will
certainly make oscillation much more likely.
This is basically due to the fact that the op-
amps. have poor stability at low gains, and a
capacitor across the feedback loop ensures
low gains at high frequencies. For gains of
more than 20dB (x 10 gain) the i.c.s are nor-
mally quite stable, assuming that the Zobel
network is fitted and that P.S.U. decoupling is
taken care of. More than 10dB (x3 gain) is
normally acceptable, but below this there can
be problems with transient response
HEATSINK IS
INTERNALLY
CONNECTED
TO PIN 3
TCA365
}
2
3
4
5
6
“i
4
q
HEATSINK IS
INTERNALLY
CONNECTED
TO PIN &
TCA 2365
Fig. 1. Pin outs for the TCA365 and TCA2365
Practical Electronics February 1985
Supply voltage All spec's quoted at +15V for
TCA365 and +10V for TCA2365
{ In normal operation:
+18
*(or36V)
Quiescent current with amps inhibited:
{TCA2365 only)
Temperature range
Maximum O/P current
Maximum |/P voltage
VP offset voltage
/P offset current
Temperature coefficient
Input current
Input resistance
Per amplifier
(Of input offset current)
At 1kHz
J Load resistance = 470Q at
1kHz
Output voltage (Load resistance = 4.72
Slew rate
Voltage gain
/P common mode
voltage range
Common mode
rejection ratio
Supply voltage
rejection ratio
Power dissipation
Equivalent I/P noise
Inhibit input
(TCA2365 only)
Open loop, at 100Hz
Load resistance = 4700
Load resistance =4700
Gain = x 100, frequency =20Hz
Gain = x 10, frequency = 100Hz
Total for package, at 90°C
Gain = x11, I/P resistor = 10k
For i.c. turned off
For ic. turned on
*= For single supply rail operation
supply)
Fig. 2. Specifications (note different supply rails used in measuring spec's.)
Rt
(overshoot of the output on square waves) and
stability.
Hence, the circuits of Figs. 3b and 3d
should be used for low gain applications.
Although the actual voltage gains of these cir-
cuits are exactly the same as the equivalent
gains of Figs. 3a and 3c, the inclusion of Ro
actually causes the op-amps to be working in
a ‘high gain’ way. Normally, this is rather un-
desirable, since there is no apparent benefit to
the user and the amplifier has a much noisier
output voltage, but in this application the
‘pseudo gain’ helps to ensure stability at low
‘ f Rt Rt
real gains, and is to be recommended for use GAIN « = GAIN = 2
with any circuitry demanding a gain: of less FoR WR GERSET WARE FOR OPTIMUM STABILITY, MAKE
than x4, or even less than x10 to be on the rer Ro = a
safe side. RO ATR
POWER SUPPLIES
The capabilities of these op-amps to dump
several amps from the supply rails into a load
puts considerable strain on the P.S.U.s used.
The best general guidance that can be given is
to consider the devices as audio power am-
plifiers, and to use the same constraints about
removing earth loops, supply decoupling,
keeping inputs away from outputs, etc. As
with audio power amplifiers, the TCA365 and
TCA2365 will overheat very rapidly when os-
cillating at very high frequencies, so any de-
bugging of stability problems should be done
very rapidly, and for short periods only!
Specifically, it is good practice to take the OV GAIN = ae cain RL* RG
connection to the feedback resistor, input fe
resistors, input decoupling, etc, as appropriate, _ REAR Rt
to the power supply as a separate connection a £4) Rr ae 7 No = ag
from the load, Zobel network, etc, to isolate
the input as far as possible from the output..In
all cases, each power op-amp should have Fig. 3. Basic power op-amp circuits
uP +VE
Rg
ov
FOR MINIMUM OFFSET, MAKE FOR OPTIMUM STABILITY, MAKE
Fig. 3c. Non-inverting, high gain Fig. 3d. Non-inverting, low gain
Practical Electronics February 1985 29
(ADJUST TO SUIT SENSOR |
(SENSOR PRESET)
22k
ite
10k
$
HYSTERESIS )
; {SETS
TCA36S
lie
210A 2365
THRESHOLD,
PRESET
+VE SUPPLY
220g IN4002
RELAY
2
IN&OG2
TCAIES TCA 2365
ov
Fig. 4. Simple sensor detector and switch
100uF capacitors between its supply rails and
0 volts, or simply across its rails in the case of
single supply systems.
When inductive loads are to be driven,
diodes should be connected between the op-
amp output and the supply rails, as shown in
Figs. 4, 6 and 8. This protects the op-amp’s
driver transistors from the huge back e.m.f.
spikes generated when inductive loads are sud-
denly turned off.
APPLICATIONS
The uses for these i.c.s fall mostly into the
realms of control and switching. They will
amplify and drive audio signals, but not with
the fidelity that can be achieved by audio
power amplifiers specifically designed for the
task, Essentially, these i.c.s are excellent for
use on many occasions when an ordinary op-
amp simply runs out of drive capability. Some
examples of switching applications are shown
in Figs. 4, 5 and 6,
A simple sensor circuit is shown in Fig. 4
using the power op-amp as a comparator and
directly driving relays | and 2. The sensor can
be any device which varies in resistance in
MARK/ SPACE RATIO
2k? TN A1GB
TCA36S
OR
VW Toaz3B5 + VE SUPPLY
100n
FOR TCA365
2k7 FREQUENCY 220n
FOR TCA2365
ov
Fig. 5. Power pulse generator
30.
proportion to a required effect. For example, a
light dependent resistor (e.g. ORP12) or ther-
mistor would allow the sensing of light level or
temperature respectively. The sensor preset
scales the voltage range produced by the sen-
sor at the op-amps non-inverting input, while
the threshold preset alters the level at which
the op-amp changes state. Ry, provides some
hysteresis to stop the op-amp ‘hunting’ or
‘chattering’ when the sensed value is just on
the threshold point. :
A square wave oscillator is formed by the
power op-amp in Fig. 5, The mark/space ratio
potentiometer adjusts the charge and dis-
charge paths for Cy such that their sum is
always constant (i.e. the frequency does not
vary) but the mark/space ratio can be ad-
justed over a wide range. The frequency itself
is set by a combination of the value of Cy and
the setting of the 100k ‘frequency’ poten-
tiometer. This circuit is capable, by virtue of
the power op-amp, of driving pulses of several
100n FOR TCA3E5
220n FOR TCAZ365
INPUTS
amps into any suitable load. Output diodes
should be added, as shown in the other cir-
cuits, if the load is to be an inductive one.
DIFFERENTIAL DRIVING
Finally, Fig. 6 shows two power op-amps
driving a d.c. motor in a ‘bridge’, or differen-
tial drive mode. This allows the direction of
rotation of the motor to be changed. IC] and
IC2 are arranged as comparators with Ry and
Ry setting the threshold voitage V7, and are
designed to be driven by logic signals A and B.
If both inputs A and B are at a low level (logic
0), both sides of the motor will be at a low
level (OV). If both inputs are high (logic 1),
then both sides of the motor will be at a high
level, near to the +ve supply rail. In both these
cases the motor will not run, since there is no
differential voltage across it—both terminals
of the motor are at the same voltage.
However, if one input is high, and the other
low, the motor will run in one direction or the
other. Normally, Vz should be set to a suitable
level for the logic family which is used to con-
trol IC1 and IC2; ideally, Ry and Ry should
be taken from the logic’s power supplies, not
the +ve power supply as shown, to ensure ac-
curacy of the threshold voltage.
The TCA365 and TCA2365 are ideal for
usé in controlling motors, relays, magnetic
valves, and solenoids. They can also make a
good basis for the design of regulated power
supplies. Their current limiting makes them es-
pecially suitable for driving filament lamps
and other unusual loads, and their op-amp
configuration makes for easy interfacing of
these loads with both analogue and digital cir-
cuitry. When stability is taken into considera-
tion these are easy and effective to use, and
provide an economic solution to many power
driving problems. Both i.c.s are available from
Electrovalue, 28 St. Jude’s Road, Englefield
Green, Egham, Surrey.
*VE SUPPLY
IN4002 INLOOZ
MOTOR
Ae
100n FOR TCA365
220m FOR TCA2365
ICi, 102 = Tca265 oR 2 TeA2365
PE6CM
Leta jeret
co [e | storren
To_[-+ [Aw cuscawise
Cpe psror
Fig. 6. Bi-directional motor control
Practical Electronics February 1985
MICROPROCESSOR CONTROLLED >
DC MOTOR DRIVERS
AST month we looked at a timer circuit
triggered by a microprocessor. This month
we have another microprocessor based pro-
ject, to allow the analogue driving of d.c.
motors, Again, the circuit assumes the use of
the Z80 microprocessor, although it is very
easily adaptable for other devices. The cir-
cuitry is shown in Figs. 7 and 8 and the Vero-
board layout in Fig. 9. The circuit consists of
two separate sections; the decoder, and the
driver. Up to 8 drivers can be operated by one
common decoder,
The address decoding is done in a similar
way to last month’s project. IC 1 compares the
most significant nibble (4 bits) of the 8 bit port
address with the settings of S1 to S4. Each
switch is turned off to correspond to a logic 1,
and on for a logic 0. The comparison is only
enabled when both TORQ and WR are at logic
0 (determined by IC5c), corresponding to the
microprocessor performing an I/O write in-
struction. [C2 is a 3-to-8 line decoder used for
the least significant nibble of the port address.
Address line A3 must be held at logic 0 and
the other 3 lines then provide the address of
the driver circuit required, The least significant
nibble of the driver address will therefore be
to 7, as determined by the latch output of [C2
used, The outputs of [C2 are inverted, and are
wired to the CATCH inputs of any required
driver circuits. Hence, if one particular driver
circuit LATCH input was connected to
LATCH output 3 of IC2, switch S1 was off
and switches $2 to $4 were all on, that driver
circuit would respond to address 13H (i.e. 19
in decimal). Driver circuits can thus be
provided at port addresses 00H to 07H, 19H
to 17H, 20H to 27H, 39 to 37H, ete.
IC3, with associated components, provides
a 5 volt regulated supply. This can be omitted
if the microcomputer’s own 5V supply is to be
used to power the logic supply to the circuitry.
1C4 provides a reference voltage which tracks
the Vp power supply. This reference will be
approximately 1-8V for a 12V supply.
THE DRIVER CIRCUITRY
IC6 is a digital to analogue (D/A) conyerter
with a built-in data latch, which connects to
the microprocessor's data bus. Data is latched
in by the required output of IC2. IC7 is a
TCA2365 dual power op-amp which drives
the motor differentially to provide both
forward and reverse control from a single sup-
ply voltage. (It’s based on an analogue version
of the differential driver in Fig. 6). IC7a am-
plifies the output of IC6 and provides the
positive output phase, while IC7b inverts this
positive drive signal about a half-rail reference
voltage set by R17, R18 and VR2, and
provides the negative output phase.
When the output of [C6 is at OV, pin 1 of
IC7a is near to OV and pin 9 of IC7b is near
to the +ve supply. When the output of IC6 is
at Vrep, the reverse is true. When IC6’s output
is at half V;e,, both power op-amp outputs are
at half the supply rail and the motor is
stationary. Presets VRI and VR2 alter the
gain and offset of the output voltages, and D1,
Practical Electronics February 1985
«Vp
t D2
1N4O02
*VpO
HP Ic3 oP
+VE POWER 7805
INPUT on
{+8V TO +201
OVp
OV POWER
FROM INPUT
MICRO- | Toro go
PROCESSOR
CONTROL
LINES
WRO
, FROM
MICRO ~
PROCESSOR
ADDRESS
Bus
IC5b
IC5 = 74LS02
== 03
IN4002
R19
270
09
tf
OVE
TO PIN 14
SET 1/0
ADDRESS,
HIGH
NIBBLE
LATCH
OUTPUTS
ON/C te t= Tr"
3 c cr
TO PIN 7 100n 1001 25¥V
6 Its osc "Poise Ti
Ov,
Fig. 7. Decoder circuit
a ‘*Bi-Colour’ Le.d., glows green for forward
direction, red for reverse and turns off in the
stationary condition.
The use of an 8 bit D/A converter allows
127 forward speeds and 127 in reverse,
although it is unlikely that the electric motor in
use will operate all the way down to OV. Note
that OV,, the logic zero volts supply, and the
power zero volts OVp, have been wired back
separately to the power input area around IC3
to help ensure stability and an absence of
noise problems with the logic supplies.
GREAT CARE must be taken when wiring
up the circuitry and assembling components
on the Veroboard, and tests should be done
prior to connecting to the computer as far as
possible. I can assure you, from practical ex-
perience, that connecting +12V to the data
lines by accident will certainly cause some in-
teresting permanent changes to the way that
your computer operates!
Vref, and the half rail reference to IC7 pin 7,
are both derived from the +Vp rail to ensure
tracking of the motor drive outputs if Vp
varies at all under different load conditions.
This circuit has successfully been used to
control a 12 volt model train set by microcom-
puter. The only programming requirement is
to output the relevant motor speed values to
the port or ports in question. Hence, for a port
address of 24 (Hex) for example, the assem-
bled Z80 machine code for full speed forwards
31
{*8¥ TO+20¥}
R12
470
db.
QUTPUT
RED 70 MOTOR
= ij
D4
Bl -COLOUR
LEO
% = CB BCS SHOULD BE 100n
(F TCA365°S ARE USED,
\ +*5V0 ub
(FROM IC3) c6
100n
19 DISC
DIO OVE
6 VREF
DEO © (FROM IC4)
FROM |05Q
MICRO- log DATA r RE
PROCESSOR 03 Wes
DATA “ 10k
BUS |D2 06 1C6 .
7
BIO ZN&28E -N
b00
LATCH
7 [FROM IC2 PIN)
; ov, O
z [FROM IC3)
‘ sis
RIT
10k
7
R18
Bk2 IC7 = TCA2365
en
100n
ibaa
a |
ov,
could look like this:
3E 00 LD A,#FFH
; Put required value in Accumulator.
D324 OUT (24H),A
; Output the Accumulator to the port.
In BASIC, the simple instruction OUT 36,-
ZSEr Rew re ame a o>
Fig. 9. Veroboard layout
255 would suffice. (36 is the decimal
equivalent of 24H, the port address, and 255
the decimal equivalent of ®FFH.) For full
speed reverse use the value 00H (@ decimal),
for stopped use 97FH (127 decimal), and for
slower speeds use values in between. Finally,
Fig. 8. Driver circuit
don’t forget the heatsink on IC7! The i.c. has
been placed at the edge of the board to allow
for this. The resulting motor speed control
provides a simple illustration of a typical use
for power op-amps, either as two single
devices or a pair as used specifically in Fig. 9.
Practical Electronics February 1985,
See
Sg AE i
AIA CENA
SE LP WCW LET
JOHN M. H. BECKER
NY electronics enthusiast needs a signal generator and
frequency meter nearly as much as a soldering iron and
multimeter. The last two should be part of anyone's
workshop but the degree of enthusiasm does not necessarily
warrant the expense of highly accurate generators and coun-
‘ters. Often only an indication of approximate frequencies is
required, together with a unit that makes readily controllable .
sounds with suitably shaped waveforms.
SWEEP GEN
CONTROL
SELECT
SLOPE RANGE
SELECT SELECT
This unit has been designed as a reasonable quality,
moderate cost, dual purpose unit suitable for average and
addicted constructors alike. It produces well shaped
waveforms of frequencies ranging from 2Hz to 78kHz in four
tunable ranges, and includes automatic ramp control of a
frequency sweep, both upwards and downwards. Ad-
ditionally it Includes a frequency to voltage converter that
can be coupled to an ordinary multimeter, or digital
voltmeter to give a direct read out of the approximate fre-
quency being generated, or fed in from an external source. It
is intended for use with an existing power supply or, for short
periods with batteries, from 9V up to 18V dc. Provision has
also been made to mount discrete power supply compo-
nents directly onto the p.c.b. so that the unit can be fully
independent of other equipment.
Photograph illustrating the
external assembly of the
Signal Generator and F-V
Converter
Fig. 1. Block diagram of the Signal Generator and Frequency-Voltage Converter
GENERATOR CHIP
An XR2206 function generator chip has been chosen in
preferance to the normally selected type 8038 as it has a
greater variety of waveforms available, together with a wider
sweep range on each selected setting. The oscillograms
show the wide range of waveforms available. The basic fre-
quency range-is selected by $1, bringing in the desired fre-
quency setting capacitor C4—C7. The frequency generated
SELECT
o f OUT
© METER
can then be controlled by either a varying voltage or a vary-
ing current. For normal manual! selection of the desired fre-
quency, current control is used, and is relative to the
resistance of the total of VR4, VR5 and R8. In this mode
VR5 is taken directly to ground by S5. As the resistance of
these potentiometers decreases, so the output frequency
rises in relation to the formula: f = 1/{R x (C/1000)) x
1000, where C is the value of the selected capacitor C4 to
C7 in microfarads, and R is the total resistance in circuit with
pin 8 of IC2.
VR4 provides coarse tuning of the frequency, and VR5
fine tuning. The maximum resistance range that is permissi-
ble with IC2 is from 1K to 2M, though is limited to a max-
imum of about 1M in this unit. This atlows a reasonable
overlap between the switched ranges, without making the
SPECIFICATION...
TOLERANCE
The figures quoted refer to those obtained on the
prototype and may vary slightly in other units in accor-
dance with normal component tolerance factors.
FREQUENCY TO VOLTAGE CONVERTER
Good linearity from 200Hz to 30kHz directly
readable on a standard multimeter or digital voltmeter.
Accessible internally and externally.
+VE 52A—C ARE GANGED POWER LINE +VE
S4A-B ARE GANGED 9¥ TO 18¥ OC
S64-B ARE GANGEO
ICt = TLOBS
R? PING =+VE, PINT] =0V
4k? or
IN4148
02
WRI
00k
c20 cg ce
On
C18 USED IN
DIFFERENT POSITION 777
WITH OPTIONAL PSU
EXTERNAL FREQUENCY
INPUT TO F-¥ CONVERTER
$10 FSI
OPTIONAL POWER SUPPLY
{NOT INTEGRAL PART OF PROJECT)
FREQUENCY GENERATOR a
Basic switched frequency ranges = (1) 2Hz to
81Hz, (2) 20Hz to 851Hz, (3) 200Hz to 8400Hz, (4)
1970Hz to 78800Hz. Coarse and fine tuning of selec-
ted. frequency range. Switch selected waveforms— —-
sine, triangle, square, ramp, pulse, and variations (see
‘photographs 1—6). Sweep modulator—tising and fall-
ing ramps, switch selectable, rate 6 to 40 cycles per —
minute. Frequency outputs—switched, buffered or un-.
buffered-via amplitude contro! from nit to 5V peak’ to | *
peak. Fixed OV/+5V amplitude square wave derived »
from internal oscillator or external source-up to about... |
80kHz. Four switched reference frequencies,
60
1
© S2A RABE]
+VE
1C3
9400CT
A ND
22u
Fig. 2. Complete circuit diagram of the Signal Generator and Frequency-Voltage Converter
Practical Electronics February 1985
35
fine tuning too coarse. Switching in VR3 by S6 instead of
VR4 and 5, a preset reference frequency can be selected.
The-i.c. contains its own current controlled amplifier and the
amplitude of the signal generated as seen at pin 2, is pre-
settable by VR6. This controls the sine, triangle, and ramp
waveform maximum levels. The squarewave however is
derived from a different section and is at approximately full
line level amplitude as determined by the current through
the load resistor R12. The shape of the triangle and ramp
waveforms is predetermined within |C2 itself. For sine wave
related waveforms, shaping is preset by VR8, and the sym-
metry trimmed by VR7. :
WAVEFORM SELECTION
Three basic waveform selections can be chosen with $2.
With S8 open, the choice is sine, triangle and square. In
position 1 (sinewave) the output comes from pin 2 of IC2,
and VRB is in circuit, controlling the sine shape. In position
2, (triangle wave), the output again is from pin 2, but at a
level approximately twice that of the sine wave and VRB is
out of circuit. In position three (squarewave), the output is
taken from pin 11 1C2. With S8 closed the squarewave is
directly fed to the contro! pin 9, and internal circuitry of the
chip is automatically switched by it to produce ramp related
waveforms in the first two positions of S2. In this mode the
frequency of oscillation now becomes affected by the value
of R9 and the formula changes to: f = (2/C/1000) x (1/R) x
1000, using the same parameters as before. Effectively this
means that the frequency with S8 closed will be approx-
imately twice that with it open. In position $2, looking at the
inverted output of IC1C, the rising ramp is sine shaped,
followed by the steep drop. In position 2 the rising ramp is
linear, again followed by a steep drop. Study of the second
formula though will show that the steepness of the drop is
related to values of RQ, and the controlling resistance on pin
8. As the two resistances approach equality, so the
steepness lessens, and a falling ramp also develops. The
best ramps are thus created with the resistance on pin 8 at
the greater end of the scale. In position 3 the output is again
from pin 11, but consists of a mark-space pulse, the duty
cycle of which determined by the formula: RA/(RA + RB),
where RA is the resistance on pin 7, and RB that on pin 8.
The negative going pulse length is moderately constant
throughout the range for the same capacitance selection.
The mark-space factor is also reflected in the shape of the
ramps with a flattening of the apex, but is really only
noticeable at small values of capacitance.
Sine (normal) $2,
Sine (ramp) $2,
36
Triangle (normal) S22
Triangle (ramp) $22
OUTPUT ROUTING
In most instances it is preferable for the amplitude of the:
different waveforms seen at the final output to be roughly
equal. As previously seen there is an inherent level difference
between the three main waveform ranges, IC1C is thus in-
cluded to even these out. The gain of this stage is of course
dependent upon the relationship of the total input resistance
to the value of the feedback resistor R16. The choice of
resistors R13-R15 ensures a reasonable match of the levels.
The inverted phase output from IC1C is decoupled by C9,
taken via S3 to the level control VR9 and then to the output
_ via S9. However the frequency pass range of IC1C is less
than that of the range available from IC2. For normal audio
applications, the frequency response of !C1C is sufficiently
adequate, but distortion becomes more prevalent as the
frequency rises above about 30kHz, as shown in the os-
cillograms,
Additionally the loading of C9 causes square wave distor-
tion at lower frequencies. S3 is thus included to bypass ICTC
so that the output is unbuffered allowing the full range of
IC2 to be used. Note though that the unbuffered output also
contains a d.c. bias that is approximately half fine level with
S2 in positions 1 & 2, and that the phase is inverted.
SWEEP OSCILLATOR
When testing out some circuits it is sometimes preferable
for the frequency range to be swept upward or downwards
at a controlled automatic rate rather than by manual control
of a potentiometer. The ramp generating circuit around IC1A
& IC1B provides this control. The frequency range of ramp
generation is determined by C1 with larger values giving
slower rates. VR1 provides the tuning of the sweep rate
setting. The direction of the ramp produced is governed by
the direction of voltage flow through D1 and D2. $4 selects
the diode routing, and reverses the polarity of the controlling
voltage through VR1 in relation to the reference level at C2.
The changing d.c. voltage produced by the ramp controls 1C2
via VR2, S9, VR4 and VR5. In this mode voltage control of
IC2 is being employed in addition to current control, For
correct operation of IC2 the voltage sweep seen at the wiper
of VR2 must lie below 3V, above this and the oscillator of
IC2 will cease. VR2 thus needs adjustment to keep the
sweep voltage within this range. The frequency control
range provided by the varying voltage is less than that
produced in the manual mode, and VR4 and VR5 are used to
select the desired band width.
Square (normal) S23
Square (pulse) $2,
Practical Electronics February 1985
Sine (high freq) $2,
FREQUENCY TO VOLTAGE CONVERTER
IC3 performs the f-to-v conversion, producing an output
voltage that, within the range, is related linearly to the fre-
quency fed in. The range available is determined by the gain
given to the feedback via R27 and VR11, with a slew rate
and ripple reduction level set by C14. High values for C14
will give reduced ripple for lower frequency signals, but will
increase the time taken for the voltage to stabilise when the
frequency is changed. The relative minimum output voltage
in the absence of an input frequency should be as close to
zero as can be set by the bias control VR10. VR11 is used to
set the maximum range. The output voltage is referenced to
an intermediate level of about 5V as set by Zener diode D5.
The negative lead of the meter used to monitor the voltage is
taken to this level, and the positive lead to the output from
pin 12, IC3. If the meter negative lead were to be taken to
the normal OV or ground line then the reading would also
contain the reference voltage of 5V and inaccurate readings
would result. For stable operation of the conversion, the in-
put frequency seen at pin 11, IC3 should be at a constant
level of about 1V p-p. To maintain this amplitude even for
low level input signals, the signals are taken via the gain
stage IC1D for external signals which gives an amplification
of about, a little over 100. The signal is then attenuated to
the optimum level by diodes D3 & D4. For internal fre-
quency reading, the signa! is taken direct from the output
of IC1C, and similarly attenuated. S7 selects the choice of
internal or external frequency monitoring. In addition to
producing a frequency related voltage, iC3 also produces a
square wave output of 5V amplitude. There ts a slight time
lag between the edges of the input frequency and of the out-
COMPONENTS...
RESISTORS
R1,R7,R15,Rt7,R22,
R25-R27 100k (8 off)
R2,R3,R18,R19 4k7 (4 off}
R4 470
R5 47k ict
R6 18k I1C2
R8,R10,R11,R12,R21 1C3
R28-R30 10k (8 off} D1~D4
R9,R24,R31 1k (3 off) D5
R13 680k
R14 180k
R16 390k
R20 1M2
R23 510k
All resistors {W +5%
CAPACITORS
C1-C3,C8-C10,C17 22, 10V elect. (7 off)
C4 1p, 63V elect.
C5,C11,C12,C14-C16,
C79,C20 100n polyester (8 off)
C6 10n polystyrene
C? 1n polystyrene
C13
ci8
VR1
VR2
VR4
VRS
VR6
VR8
VRS
VR10
VR11
Practical Electronics February 1985
Triangle (high freq) S2,
56p polystyrene
470n, 25V elect.
SEMICONDUCTORS.
Ramp (high freq) $2,
put 5V squarewave. S9 can switch in this frequency output
in place of that produced directly by 1C2. This means that an
external frequency of indifferent level and shape can be
converted for controlling circuits that require a 5 volt
squarewave. The external loading permissible though is
limited by the value of R30 and too great a toad will reduce
the voltage.
POWER SUPPLY
Most enthusiasts probably already have power supplies in
their workshop capable of driving this unit, and so a separate
one is not included. The minimum voltage requirement is 9V,
and the maximum permissible +18V as dictated by the limits
of IC3. The current drawn is about 30mA, up to 20mA of
which is consumed by IC2. This current is a bit too high for
the unit to be powered for long periods from a battery sup-
ply, though one could be used briefly in an emergency. Alter-
natively a battery eliminator might be suitable, providing it
can tolerate the current without the voltage dropping below
QV and that the ripple content is negligible. The printed cir-
cuit board though includes positions for the mounting of the
rectifier and voltage regulator as shown in the suggested op-
tional 12V power supply circuit. This supply was not used in
the prototype and is not regarded as an integral part of the
project. The transformer should be bolted to the metal box,
and normal mains electricity safety precautions observed.
Note that with this suggested power supply C18 has its
positive end connected to a different track position, and it
may be necessary to mount it vertically rather than horizon-
tally. The use of a heat sink with the regulator i.c, should not
be necessary.
MISCELLANEOUS
P.c.b, and p.c.b. clips (4 off}
Round knobs:(6 off)
Le. sockets, 16 pin, 14 pirn(2 off)
Jack socket, 3-Smar 7
VR3,VR7
~ TLO84 (quad op-amp)
XR2206
9400CT (R.S. 307 070)
1N4148 (4 off}
5V1 Zener, 400mW
POTENTIOMETERS
~ 100k mono rota
10k skeleton
25k skeleton (2 off)
1M mono rota »
25k mono rota
50k skeleton
500 skeleton
10k mono rote
100k skéleton-:
250k skeleton
Jack sockets mono {2 off).
Box and riibber feet»...
Meter terminals. (2 aff
SWITCHES
$1,82
$3,S5,S7-S9
Sage ce
3P4W (2 off)
SPDT (5 off)
» DPDT (2 off)
codeweenrs Hote me
A complete kit of parts is available from:
Phonosonies, 8 Finucane Drive, Or-
pington; Kent BRS 4ED. Price £54.00,
inciusive of VAT, Post end packing
£1-00.
37
ASSEMBLY
After the straight forward component assembly and sub-
sequent joint checking procedure has been carried out, wir-
ing should be commenced in a methodical fashion, ticking
off each wire on the wiring diagram as connections are
made. First connect up all the panel controls between them-
selves. Secondly wire up to all-the p.c.b. points closest to the
front panel. Finally connect up the rear and remaining con-
nection points. These latter wires should preferably be
brought under or round the edges of the p.c.b. Taking them
over makes the wiring untidy. Keep the wiring short, but
long enough for turning the board over for examination
without over straining the connections {too much flexing of
taut wires can cause breaking at the joins). The prototype
has the meter terminals on the front, but with hindsight,
mounting them on the back would be better.
The regulator |C4, and the rectifier REC1, are part of the
optional power supply, together with the transformer T1.
These components may be omitted if not required as the
unit will run quite efficiently from any 9V battery.
SETTING UP
A fair selection of presets has been included to enable the
maximum accuracy to be obtained throughout the unit. The
only really critical one is VR2, as the sweep control may not
operate if this is incorrectly set. Inadequate setting of the
others will only cause lack of linearity. If an oscilloscope is
not available intelligent decisions will need to be made,
listening to the sounds while adjusting the presets.
First, S1 position 1 (lowest freq), S2 position 1 (sine), S3
on (buffered output), S4 either way, S5 off (sweep off), S6
off (manual freq control), S7 off (internal f-v), S8 off (stan-
dard waveforms), S9 off {vco output), VR1-3 midway,
VR4—5 max clockwise (highest freq), VR6-8 midway, VR9
max, VR10—11 midway. Switch on and check that pin 3
IC1A, pin 3 IC2 are at approximately half line voltage, and
that the positive end of D5 is at about 5V. Plug in to normal
amplifier. If no sound is heard then check the wiring, and
that the switches are wired the correct way up. Assuming
that sound is heard, check that VR4 and 5 vary the fre-
quency, and that S1 changes the ranges. Switch on S6, and
adjust VR3 to the desired fixed frequency on the second
range of S1. In the author's unit this was set for approx-
imately 440Hz. Check that S8 brings in the ramp associated
38
% DOTTED LINE SHOWS CONNECTION
OF CIB IF OPTIONAL PSU IS USED
## 1C4 AND RECTIFIER ARE PART OF OPTIONAL
PSU ANO NOT INTEGRAL PART OF PROJECT.
- VR2
ie
G8 Caede
ct
D
ny
2
3
7
+
al
w
+
oO
©
e
is)
©
e
]
ied bh
A
a} io
+
See
5
w
?
(PES? 3A)
Fig. 3. (Above) Showing the
p.c.b. design. and component
layout of the Signal
Generator. (Left) Photograph
illustrating the internal view
of the chassis assembly
showing the switch, sockets
and p.c.b. layout
Practical Electronics February 1985
xT,
INPUT OUTPUT
»
$1 PIN 9 ONLY USED AS CONVENIENT OV TAG
PCB PINS 29 & 30 GO TO TRANSFORMER 12¥
TAGS IF USED, IN WHICH CASE PCB PIN 27
IS NOT USED.
nv 17 33
INT /EXT
F-¥
Fig. 4. Wiring diagram of the Signal Generator
ranges of waveforms, then switch back to the normal range,
and to squarewave. Monitor the output jack socket and note
the squarewave amplitude level. Switch to triangle wave and
adjust VR6 until the amplitude is similar without flattening
of the waveform peaks. Switch to sine wave and adjust VR8
for the best sine shape, then VR7 for the best symmetry. If
necessary readjust VR8.
Switch S3 to bypass and check all waveforms can. be
switched in, though as previously stated they wilt be at
widely varying levels. Switch S6 back to manual coftrol,
then S5 to sweep control. Check that a ramp waveform ap-
pears at pin 1 IC1A, and that switching S4 varies the direc-
tion, also that VR1 varies the rate with the ramp in both
directions. Return scope probe to the output jack socket, and
adjust VR2 for the smoothest sweep response. If the wiper is
too far to the OV end the generator frequency will dwell at
the high end, with it too far towards IC1 the generator will
cut out at the low frequency end. With careful adjustment a
smoothly varying rising or falling sweep range can be set.
Switch off the sweep control and set a frequency output
from !C2 of precisely 1O0kHz. Check that an attenuated ver-
sion of this frequency reaches pin 11 IC3 via S7. With the
external input jack socket grounded (as it will be without a
jack plug in), switch S7 to external. Connect a multimeter
across the meter output terminals and adjust VR10 for a
reading of exactly zero volts. Start off with a meter range of __
Practical Electronics February 1985
about 5 volts, then after the initial adjustment has been
made the. meter can be switched to its lowest range for
greater precision. Set meter to a range for monitoring exac-
tly 1V. Switch S7 back so that the 10kHz signal reaches IC3.
Very carefully adjust VR11 until a precise voltage of 1 volt is
obtained.
The maximum frequency that can be read will be about
30kHz to 35kHz, beyond this IC3 will fail to respond and
show. @ constant reading of around 3-5V. Applying a 1kHz
signal should produce approximately 0-1V:: Tracking
downwards in frequency, linearity will be roughly maintained
until about 200 Hz or so, depending on the accuracy of the
setting of VR10 and VR11. Once the boundary extremes
have been established, a direct reading of frequency can be
taken from the multimeter by converting the voltage into the
readily calculable frequency. Thus if 1V = 10kHz, 3V =
30kHz, O-5V = 5kHz, 0-O5V = 500Hz, etc. Check that
an external frequency can be monitored in the same way,
then finally that. a squarewave output of about 5V is
available from pin 8 to IC2.
After setting up the f-v converter the frequency controls of
the unit can be calibrated and control legends applied to the
panel, using a rub down lettering like letraset or similar, then
coating them with a suitable spray protector. If care has
been taken in the setting up, the end result will be a mar-
vellously versatile dual purpose new unit for the workshop. %&
RING MASTER
Perhaps the most interesting development
during the past few months has been the visual
detection of the ring-system surrounding
Uranus. The rings were first found indirectly,
because they produced a series of occultations
of a faint star; subsequently, D. A. Allen and
J. Crawford, at Siding Spring Observatory in
Australia, photographed them in infra-red.
Studies of them have now been carried out
by Richard Terrile and Bradford Smith, using
the 2-5-metre reflector at the Las Campanas
Observatory in Chile together with a highly
sensitive CCD or Charge-Coupled Device.
The ring-system is clearly shown, together
with all five known satellites—Miranda, Ariel,
Umbriel, Titania and Oberon. The pictures
show the great power of the CCD, which is at
least thirty times more sensitive than any
photographic plate.
The rings of Uranus are quite unlike those
of Saturn. Terrile and Smith find that their
albedo or reflecting power is only about 2 per
cent, so that they are blacker than coal-dust.
They are also narrow; there are at least eight
rings, not all of which are perfectly circular,
and their composition is unknown.
If all goes well, we should learn more about
them-in January 1986, when the Voyager 2
spacecraft makes its pass of Uranus.
Meanwhile, there is speculation about the
possibility of a ring-system round the outer-
most giant, Neptune, but the presence there of
a large retrograde satellite (Triton) may have
prevented any rings from being formed.
Again, we pin our hopes on Voyager 2, which
will rendezvous with Neptune in the late
summer of 1989.
HALLEY’S COMET
Halley’s Comet is, of course, still much too
faint to be detected except with very powerful
instruments, but recent studies show that it
may prove to be somewhat brighter than had
been expected. Unfortunately, this does not
mean that it will be a brilliant spectacle, as it
has been on many past returns.
It should become a naked-eye object at the
end of 1985, before perihelion passage on 9
February, 1986, but British observers will
need clear skies. When at its best, after
perihelion, the comet will be in the far south—
well placed for Australians and South
Africans, but not for Europeans, who will not
see it at all until it has faded considerably.
NOVA CYGNI—A NEW LOOK
On the evening of 29 August, 1975, | went
into my observatory to make some routine ob-
servations of variable stars. When I looked up
at the familiar constellation of Cygnus, I had a
surprise. There, shining down unmistakably,
was a bright star which had certainly not been
there on the previous night. | estimated its
magnitude as 2-4, slightly fainter than
Gamma Cygni, the central star of the “cross”
of Cygnus.
THE SKY THIS MONTH
Having satisfied myself that it really was a
new star or nova, | made a telephone call to
the observatory at Herstmonceux. I was, of
course, fairly sure that the star had already
been reported—and so it proved; it had been
discovered some hours earlier by Kentaro
Osada in Japan, before darkness fell over
England. I imagine that I was about sixtieth in
the list of independent discoverers; the star
could not possibly have been overlooked by
anyone with more than a rudimentary
knowledge of the constellations.
The most remarkable fact about Nova
. Cygni was that it brightened up by at least
nineteen magnitudes in only a’few hours. This
was a record, both for amplitude and for
speed. Its decline was also unusually quick. I
estimated its magnitude as 1-8 on 30 August,
so that it was then much the brightest star in
the constellation apart from Deneb; but it had
dropped to below 3 by 1 September, below 5
by 4 September, and faded below naked-eye
visibility by 7 September, when I saw it as
fiery red—in fact, as red as any star I have
ever seen. Within a few months it had become
too faint to be observed except with powerful
telescopes.
Apart from Nova Cygni, only three novae
seen since 1930 have attained the first
magnitude: DQ Herculis (1934), CP Lacertae
(1936) and CP Puppis (1942), though others
have become visible with the naked eye-—
notably HR Delphini, which was discovered in
1967 by the well-known English amateur
George Alcock and had a very prolonged
maximum of around the fourth magnitude. It
is still above magnitude 13, and probably will
not fade much further, as this was also its pre-
outburst magnitude,
However, the exceptional behaviour of
Nova Cygni has led to particularly detailed
studies of it, and efforts have been made to
detect a cloud of débris round it. This’ has now
been a successful operation,
40
Winter skies are always glorious, thanks to the
presence of Orian, the Hunter, and his magnificent
retinue, but at the moment the dearth of bright planets
continues—apart from Venus, which is at its very best in
the evenings. Mercury is, in theory, a morning object,
and may indeed be glimpsed just before sunrise, but it is
well south of the celestial equator, so that European ob-
servers are unfavourably placed.
Mars may be seen in the south-west during the early
evening, and moves from Aquarius into Pisces by the end
of the month, but its magnitude is now only 1.2, and no
telescope will show much upon its surface. Saturn, in
Libra, rises well before the Sun, but is low down and by
no means prominent, while Jupiter passes through con-
junction on 14 January and is therefore out of view
altogether. There are no eclipses this month, and no
bright comets are expected. The Moon js full on 7
January, and new on the 21st.
During winter evenings the brilliant yellow star
Capella is almost overhead (a position occupied by the P
equally brilliant Vega during evenings in summer}. Close
beside Capella lie the three fainter stars making up a
triangle. They have been nicknamed the Haedi or ‘Kids’,
and two of them are very remarkable objects.
Epsilon Aurigae, at the apex of the triangle, is an ex-
tremely luminous supergiant, at least 60,000 times as
powerful as the Sun. Every 27 years it fades down by
almost a magnitude, not because it is intrinsically
variable but because it is being eclipsed by a companion
which has never been seen at ail.
The nature of the invisible secondary is still a matter
for debate. It was once believed to be a very young star,
not yet hot enough to shine; there were also suggestions
that it might be a black hoje, but it now seems more
likely that it is a relatively small, hot star with an
associated extensive shell of material. The last eclipse
ended in 1984, so that for more than two decades
nothing further will be happening.
Look at the ‘Kids’ and you will see that Epsilon is now
the brightest member of the trio. The faintest, Sadatoni
or Zeta Aurigae, is also an eclipsing binary with a period
of 972 days, but we know much more about it; the
primary is a red supergiant, while the secondary is a
much smaller and hotter star.
ft is sheer coincidence that these two exceptional
eclipsing binaries lie side by side in the sky. There is no
true connection between them; Epsilon is much further
away from us than Zeta.
Practical Electronics February 1985
, Using the 82-inch reflector at the
McDonald observatory in Texas, G. and A. de
Vaucouleurs have recorded the débris un-
mistakably. The cloud shows up as an ellip-
tical blur, 3-5 x 2-5 seconds of are across,
with an integrated magnitude of about 164.
Presumably it is expanding; and if the distance
of the nova is 4,500 light-years, as seems
likely, the expansion rate is approximately 800
miles per second.
BINARY SYSTEM
According to modern theory, a nova is a
binary system, made up of an ordinary main
sequence star together with a white dwarf.
White dwarfs are stars far advanced in their
evolution; they have used up their nuclear
“fuel”, and have become very small and
almost incredibly dense.
In a nova, the white dwarf pulls material
away from its larger, less dense companion,
and there is a build-up of material around the
dwarf, leading eventually to instability and a
violent, usually short-lived outburst.
Associated debris is only to be expected, and
has been detected with many former novae—
such as Nova Persei 1901 and Nova Aquilae
1918, both of which became much more
brilliant than Nova Cygni. Indeed, for a brief
period Nova Aquilae outshone every star in
the sky with the exception of Sirius.
SUPERNOVAE
There is a fundamental difference between
ordinary novae and the much more powerful
supernovae, which are much less common.
Only four supernovae have been seen in our
Galaxy during the past thousand years; those
of 1006, 1054, 1572 and 1604—all before the
invention of the telescope, though supernovae
are so luminous that they may be detected in
external galaxies many millions of light-years
away.
From the few accounts which have come
down to us, the 1006 supernovae, in the
southern constellation of Lupus, became as
bright as the quarter-moon, while the other
three outshone Venus. The 1054 supernova
has left the gas-cloud of the Crab Nebula,
which contains a pulsar.
For many years the Crab was regarded as
unique, but recently a very similar supernova
remnant, including a pulsar, has been found in
the Large Cloud of Magellan, more than
150,000 light-years from us. Pulsars are
rapidly-spinning neutron stars which slow
down as they age. From the measured
slowing-down of the Large Cloud pulsar; it
has been estimated that the outburst occurred
about a thousand years before our present-day
view of it. ;
The rapid’ increase of Nova Cygni 1975
raised initial hopes. that it might be a
supernova-—something which astronomers
would warmly welcome, because it cannot be
said that our knowledge of supernova
mechanism is at all reliable (there seem, in-
deed, to be two quite different types of super-
novae, one of which involves the collapse of a
massive star while the other indicates the com-
plete destruction of a white dwarf).
We could learn much more if we had the
opportunity to study a relatively nearby super-
nova with modern equipment. Nova Cygni
was therefore something of a disappointment;
but it is still extremely interesting, and it will
be important to find out how the -newly-
discovered cloud of debris develops.
There is one rather sobering thought. When
we look out into the Galaxy, we are also look-
ing into the past. Nova Cygni exploded well
over four thousand years ago; the expanding
cloud we see today was produced long-before
astronomy had become a true science!
SUBSCRIPTION
RATES
U.K.
Overseas
£13.00 |
£14.00 |
COMPLETE |
ANDPOST |
THIS :
|
|
Practical Electronics
February 1985
Complete this form and
post it, with payment or
credit card authorisation to:
West Sussex RH16 3DH
DON'T MISS A VITAL COPY!
Ever been in the middle of a project
only to find the next issue sold out?
An annual subscription to PRACTICAL
ELECTRONICS solves the problem.
problems.
PRACTICAL
- BLECTRONICS 305%
ORDER FORM
Annual Subscription Rates
UK £
POST COPIES TO
Wherever you live you'll receive every
copy regularly each month. It's the
quick, practical way to solve delivery
13.00
| Overseas £14.00
(Students: deduct £1 and quote NAME
Student number)
ADORESS_. =
POSTCODE ___
lenclose my cheque/PO payable to 1PC Magazines Ltd. for £__.
Charge my credit card A/C-at quoted rate [1] vO vj
|
|
|
|
|
|
|
|
|
|
[TllLilll ili ° |
Practical Electronics a
Subscription Dept.
ORDER F ORM Oakfield ews CT]
35 Perrymount Road,
TODAY! coun fat Card valid from
Signature
41
A selection of readers’ original circuit ideas.
Why not submit your idea? Any idea published will
bl be awarded payment according to its merits,
Each idea submitted must be accompanied by a
declaration to the effect that it has been tried and E
tested, is the original work of the undersigned, and a
that it has not been offered or accepted for publica-
tion elsewhere. It should be emphasised that these
= a
designs have not been proven by us, They will at any
rate stimulate further thought.
. Articles submitted for publication should conform
to the usual practices of this journal, e.g. with regard
to abbreviations and circuit symbols, Diagrams
should be on separate sheets, not in the text,
TH ER MM J Ss TO R HE conventional faldye elven recom- 4
mended for thermistors is somewhat in-
Glass bead: thermistor accurate. Better results can be obtained us-
Epowy resin sealant TH ERMOMETER ing the arrangement shown here. The
op.amp, IC2 acts as a voltage source, with
current measured by the meter ME1. This
configuration corrects for most of the ther-
mistor’s non-linearity, An offset current
flows through RS and VR1, and sensitivity
is set by VR2 and R6. Assuming a full-
scale current of ImA, an output of
Im¥/°C is conveniently obtained at the :
output and this can be used to drive a i
chart recorder, for example. The supply
voltage is regulated by ICI, and a 3-way
switch allows the battery to be checked,
Resistor values for two types of NTC 4
thermistor are given in Table 1. :
te JA-03 is metal-sheathed (RS no. 151-120)
CALIBRATION OF METER }mA FS.D. and VA3704 (Mullard no. 2322-627-
Stainless
steel tube
CONSTRUCTION OF : ;
TEMPERATURE PROBE .. ieee
USING GLASS BEAD ,
TYPE THERMISTOR
PE 1554 P) hee resin sealant
Construction and calibration of the Thermistor Thermometer 11472) is a glass-bead device. Both have a F
resistance at 50°C, of the order of Lk7. :
; RT [ri [a2 [rs [re | R7 | Kettles and ice-buckets are unpredic- ‘
Pa-03 [7G | so] aan] 2 foc]
Pvas70~ | 10x J 647] tox | | 10% |
table gadgets! Calibration was done at 10,
50 and 90°C, using a water-bath and
laboratory-grade mercury thermometer.
The error curve is cubic, with a deviation
within 2 or 3 degrces centigrade over most :
of the range, increasing near the 0° and 4
100°C limits.
a 0 30 70 400
ERROR|+3-3 -2:0 91-7) 72-6
AT [+40 -2-5 42:2 -3-2
TABLE
2 ss aan
alt] 50 20 30
C.J. D, Catto, ;
Elsworth, ‘s
ie Cambridge, 4
4
Amv sec ¥ :
OUTPUT RS :
A *% See tablet
O O +H See text : a
YRI i
5k SET ZERO a
(20 TURN] t
rd
VR2
500 SET FS.
{20 TURN}
1 == ‘5! Positions.
gv te 1 OFF
2 BATT, CHECK
3. 0G
agt*
- t o
PROBE :
, 4
|
| ,
i 42 \ Practical Electronics February 1985
es
rf
ie
r
mage ve eRe
LOGIC
RECORDER
HE logic recorder is designed to fill the
gap between the logic probe and the
logic analyser by displaying a sequence of
eight data bits on a bank of |e.d.s. Data is
stored synchronously with the system
clock after a trigger pulse from the circuit
under test.
Data is fed into 1C4, an octal D-type
flip-flop wired as a shift register, though a
74LS164 shift register could be used in-
stead. D2 represents the most recent data.
Suppose that latch IC 1a/b is set so that
D10 is unlit. IC] pin 6 is high so IC3 is
reset, its “A’ output is low, IC Id is disabled
and ICle is enabled via IC2b. When the
ARM button is pressed the latch sets,
lighting D10 and freeing [C3 to count. On
the rising edge of IC 2c pin 8, the single-bit
counter in IC3 clocks, disabling IC l¢ and
the ARM button and enabling ICld to
clock IC4 and, via [C2d, the 3-bit section
of IC3.
After eight clock pulses from the system
under test, IC3 pin 11 goes low, resetting
the latch via CI. Provided that the ARM
button has been released the logic recorder
resets and displays the data received until
it is rearmed.
IC2a and IC2c are used as program-
mable invertérs to select the desired trigger
and clock edges. DI protects the recorder
against reverse polarity on its power
supply leads which are connected to the
test circuit,
G. Strange,
Loughborough,
Leics.
ICia ICctb
SKI 24802 241502 IC ic
a 7&LS02
b
ORG
ae ¢
RO
vo )4 ari: |
1c2 ,
74LS365
Be pies
oz} 3 |
pa iG
04
os
Practical Electronics February 1985
TRIGGER
CcLoOck
> oO
a | | [1
IN5400 J
ov
May be 1 or 0
oe
D7 D6 DS D4 D3 D2 D1
f — X X X Xx
8 2
l 1
128 64 32 16
D
x
1
Loh. fo 8 1
Xx
4
0
e.g. D2 and D4 are low ;
128 + 644+ 32+0+8+4+0+2 + 1=235
Table 1. Showing example of
1N127 command
cid 101 PIN 14
74(S02 IC2 PIN 16
3 v o 19 13
} «ths Le Le
(
ae oer
ob a be
7:
\/
Wy, peal Z 6 INPUTS
ae
rs
Ais
10142 PINTS ©
PIN B i.
[C3 PIN 10
1C 1/2 PIN?
ZX SPECTRUM
BUFFERED
INPUT PORT
(6 BIT)
OST designs for input ports use
devices such as the 74LS245 octal
transceiver, but during the current chip
shortages, these devices are either impos-
sible to obtain or very expensive. There are
other possible buffering chips, such as the
74LS126, but these are more difficult to
design a circuit layout for.
In my design I have used the 74L$365
Hex buffer/driver with gate enable
inputs—the only disadvantage is that we
now have only 6 inputs.
The circuit itself is very simple: the port
is 1/O ‘mapped by A7 and address
decoding is performed by IC1, a 74LS02.
Thus when the following control lines are
low—iORQ to indicate an input output
operation and RD to indicate the CPU
wishes to read from the I/O port indicated
by the address on the address bus, The
outputs of [C2 are put onto the databus of
the computer. C1 and C2 are decoupling
capacitors.
Since the circuit is mapped by A7 being
low, an IN127 command will give 255 if
all inputs are high {logic 1), as will be the
case if they are left unconnected. If bit 1
(D®) is low then the result will be 254.
Possible results are summarised in Table 1.
Thus the result of an IN127 command in
this case would be 235.
A. Moran,
Reading,
Berks.
43
CLOEK PULSE No o CISPLAY o
0
so" oe
VAT
z 3 [> Bs} o
a ad
cy 0
via 0
sole
7
i a [> Bolo
ee
a o
a aly °
at ° °
ce
(Above) Some possible
fault indications
(Below) Complete
circuit diagram of the
DIN Load Tester
(PEEI7E)
44
INTERPRETATION
COWMECT cite
OF Pw
Ping SHORTED ate
ho COMMER TION
To Pika er A
CROIRFE WrACS
connect
SOMMECT: ON
OF SCAEEN
2—— TOU
<All
DIN LEAD
TESTER
AVE you ever wondered whether all
your problems might just be due to a
faulty DIN lead? If so, you'll no doubt
have discovered what an awkward opera-
tion it can be to test one; but here is a
simple solution.
The following design will test DIN leads
for open circuit, short circuit and wrong
connections all in one go, Whereas with a
continuity tester you need to test each pin
individually, this circuit gives an im-
mediate indication of the connections (or
lack of connections) between pins at either
end.
IC1 is a 4017, which is a decade coun-
ter with decoded outputs. It sends each pin
high, in sequence, moving on to the next
one every time it receives a clock pulse.
For a 5 pin DIN, the sixth output is con-
nected to reset and thus counts through
0,1,2,3,4 and 5. Six pulses are allowed to
enable the screen to the tested separately
(note however that this is often connected
to the middle pin). The clock used to drive
ICI is produced by a standard CMOS
oscillator based around IC2 (a 4011 quad
NAND), the NAND inputs being shorted
together as shown to act as inverters. This
i.c, type was chosen merely so that fewer
gates were left unused. A hex inverter
would serve just as well. IC3 and IC4 are
4050 hex buffers used to drive the L.e.d.s.
The clock frequency is normally about
1kHz, providing an apparently constant
display. In this mode any broken connec-
tions will immediately show up as unlit
Le.d.s. at ‘B’, By pressing switch S1 the
68k resistor R1, is disconnected and the
clock frequency reduced so that a more
detailed representation of the condition of
the DIN lead is given. Two common types
of § pin DIN lead exist; straight through
and mirror image.
STRAIGHT THROUGH—L.e.d.s at ‘A’
and ‘B’ should light in the same order,
MIRROR IMAGE——L.e.d.s at ‘A’ and 'B’
should light in opposite orders. Crossed
leads can thus be detected by incorrect
orders at ‘B’,
Any two Le.d.s lit simultaneously at one
end in this mode indicate shorted leads at
that end.
Some sample displays are given. Using
CMOS logic. the unit can be readily
powered by a 9V (PP3) battery,
C. Walden,
Selby,
S. Yorks.
Practical Electronics February 1985
RICHARD B. H. BECKER — SYSTEM DESIGN AND MECHANICAL ENGINEERING.
TIM ORR — COMPUTER INTERFACE AND CONTROL ELECTRONICS.
ic ELECTRO-MECHANICS and the electronics of both robots
have now been explained in previous parts, and the p.c.b.s
and parts lists have been published. The actual construction and
computer control of the robots is the last thing to be considered,
beginning this month with NEPTUNE.
CONSTRUCTION
Construction starts with axis 0 and progresses upwards from
there. The hydraulic cylinders of the NEPTUNE are supplied
pre-assembled. First axis 0 cylinder is fitted to its sensor poten-
tiometer, on a bracket beneath it, and then to the base plate and
front plate (Fig. 6.1). There are support brackets to ensure a
rigid structure and there are triangular brackets with feet on
them to prevent the robot from tipping over when under load.
The top slot in the rear plate is for fitting the computer interface
leads to the edge connectors on the interface board. The slot
beneath it is for the cables leading to the power supply which fit
onto the rear plate later.
Next the top plate is fitted. Being 3mm steel this is quite
heavy but this is necessary to ensure rigidity. On top of axis 0
cylinder goes the shoulder rotating axle with axis | mounting
plate on top of it (Fig. 6.2). The weight of the arm is carried by a
large thrust ball race fitted round the axle. Through the axle
passes the plumbing/wiring harness which is secured to axis |
cylinder fitted to the mounting plate.
Fig. 6.1. (above): base plate and axis 0 cylinder
Fig. 6.2. (opposite): shoulder-rotating axle with axis 1
cylinder fitted to the mounting plate
Practical Electronics. February 1985
PART SIX
Onto mounting rings on axis | cylinder the lower arms are fit-
ted and to the upper end of these are fitted axis 2 cylinder with
the upper arms attached to it (Fig. 6.3). To the end of the upper
arms is fitted axis 3 cylinder which provides the wrist raising
function.
A pair of mounting plates is used to secure axis 4 cylinder to
the front end of axis 3 cylinder (Fig. 6.4). This applies to NEP-
TUNE II only, where it provides wrist yaw, a function immen-
sely useful when picking up objects lying flat or in other difficult
positions. It also greatly assists spraying into corners. Yaw is a
feature usually found on only.the most expensive of robots used
in industry. Axis 5, the wrist rotation cylinder fits to the axle of
axis 4 cylinder by means of a short plate. On NEPTUNE I
where there is no wrist yaw, axis 5 fits to the top of axis 3 cylin-
der. The gripper clamps onto the axle of axis 5 cylinder with a
single set screw enabling it to be rapidly changed. The NEP-
TUNES are supplied with a choice of gripper types. The robot is
now ready for plumbling and wiring.
49
mee ROBOTICS PROD
The harness is strapped onto the arms at the fixing holes with
cable ties. The position for these is marked, with dabs of paint,
on, the harness. The pipes press into ‘banjo’ fittings attached to
the cylinder ends with Delrin hollow, a a and the cables are
trimmed and soldered to the sensor potentiometers (Fig. 6.5).
The solenoid operated valves, which control the water flow,
fit with hollow bolts to a Delrin manifold (Fig. 6.6) which acts
like a printed circuit board routing the water to the correct valve.
The other end of the valves fit to bored Delrin bars to which the
restrictors and flow rate control valves also fit (Fig. 6.7).
Brackets to which the solenoid driver boards will fit are screwed
to the solenoids and ensure correct alignment for the boards
which clip onto the connector tags of the solenoids (Fig. 6.8).
This arrangement avoids soldering to the solenoids and greatly
simplifies maintenance in. that the boards can be rapidly un-
plugged from the system. Connections to the plumbing harness
Fig. 6.3. Lower and upper arms fitted
Hydraulic cylinder assembly
50
AP ore’ opi
ih
ih RarhP aa
again are made’ with “banjo” push-on fittings and hollow bolts.
After wiring up the power supply (Fig. 6.9) and connecting to’
the computer interface board (no need to connect to a computer
yet) the system is ready for commissioning. The sump of the
hydraulic power pack is filled with water and the pump is
operated with its outlet and inlet connected together with one of
the plug-in pipes. This expels the air from the pipe which, after
switching off the pump, is plugged into the manifold as the
pressure source; the return pipe is taken from the outlet of the
manifold to the sump. After pressurising the system and check-
ing that there are no Jeaks the axes are tested one at a time by
plugging in a solenoid driver board and the axis operated by
means of a potentiometer on CN401 extend-contract connector.
This will drive out most of the air. The rest of it will gradually
disappear by dissolving in the water when under pressure. It
then comes out of solution when returned to the sump.
Fig. 6.5. Plumbing and wiring loom
Practical Electronics February 1985
a meee FSR ae EE nee da
Fig. 6.6. Hydraulic contro! manifold
Fig. 6.7. Assembly showing restrictors and flow-rate control
valves
Fig. 6.8. Completed hydraulic flow-control sub-system for
all seven axes
Fig. 6.9. Power supply
Practical Electronics February 1985
All the solenoid driver boards may now be fitted (Fig. 6.10),
the computer interface board connected to the computer and the
“NEPDYN” program run. This sends and returns the chosen
axis between 2 points and generates, on the monitor screen, a
graph of error in the axis position against time. The error is the
difference between where the axis is, as measured by the ADC,
and where the axis has been told to go. With the aid of this
program the restrictors are set to achieve rapid convergence of
the send and return graphs without any overshooting.
After setting the restrictors of each axis, the pre-sets on the
solenoid driver boards are set so that sending position 0 from the
computer, by means of the “NEPTROL” program, sends each
axis to just before the end of its travel. The pre-sets on the inter-
face board are set to match the positions of the simulator (Fig.
6.11) with the positions of the robot.
The interface board is next mounted over the manifold
assembly (Fig. 6.12) which is then slid into the robot base and
bolted down. Following fitting of side plates and covers the
robot is then ready for use.
OPERATION OF THE ROBOT
Operating the robot basically consists of using POKE and
PEEK instructions sent to the robot as if it were part of the
memory of the computer. To move axis 0 of the NEPTUNE II
Fig. 6.10. All sub-assemblies completed and wired up for
testing
Fig. 6.11. NEPTUNE simulator arm
5!
te position DO, the start of the ‘memory’ is first defined. This is
jiso-the address for the most significant byte-of axis 0. The least
‘sin icated by *&’. , wane
mya?" 10 A=&FCOO : c £ ee. oe :
bonnes 20°2?A=DO DIV 16). . >. pee
30 %A+1)= LOO MO ERIE: SERN SF
The data DO can be any integer from 0 to 4095 (2"-1) because
it is‘a 12-bit control system. On ‘the NEPTUNE ‘T it is am 8-bit
system so the range is 0 to 255 (28-1) so only one byte is sent
for each axis move. For the msb the data is divided by 16 and
the remainder ignored. For the Isb this remainder is multiplied
by 16 because it is the top 4 bits of the Isb that are used.
The addresses of the axes: follow successiyely so to move axis
4to position D4 the instructions are as below.
“40 %A+8)=D4 DIV 16°,
50 KA+DAS MOD: oy 16"
Similarly for axis 5
60 XA+10)=D5 DIV 16.
70 2%A+11)=(DS MOD ‘16)* #6
The. servo system of the robot then,’ makes, the axis go to this
position with no. further computer intervention but with the
ADC the position of the axis can be followed as it is moving. To
operate the ADC it is written to at A+14, Data bit 7 is toggled
and the multiplexer axis address set up. The axis address is in
the bottom 4 data lines (see Table 1 October 1984). Axis 0 is at
address 0000. The msb i is read at A+17 and the Isb at A+ 16.
80 A+14)=128- Vere
-90 (A+IDSO OO a a 'e s
‘100 %A+14)+128 :
110 DAO=NAL17)*16+2(A+16)/16
2) * Significant byte is at the next address. On'thé BBC, POKE and »
ra." PEEK are’ Tepresented. by. ey xand! ‘hoenieginal nyinibers , ate.
Fig. 6.12. Control. and servo
electronics ready for installa-
tion in robot base
Similarly for axis 4
120 %A+14)=132
130 9(A+14)=4
140 2(A+14)=132
1S0° DA4=2(A+ 17)" 16+2(A + 16)/16
Reading the simulator. is performed similarly but at the mul- ~
tiplexer axis address 8 bits higher so to read simulator axis 0:
160 ?(A+14)=136
170 %A+14)=8
180 2(A+14)=136
190 DSO=?(A+17)*16+2(A + 16)/16
If the simulator is constantly read and the data returned to the .
robot then the robot will follow the movement of the simulator,
. NEXT MONTH: details of the assembly and use of
MENTOR.
The complete NEPTUNE 2 robot system, showing the
pump, computer and monitor, and the simulator
Practical Electronics February 1985 '
mas A
i INPUTS.
4 . .—o i]
AST month we carried out a detailed
investigation of the operation of a
‘universal shift register. This month we
«shall turn our attention to another
‘device which finds a wide range of ap-
‘plications in the digital world, the data
multiplexer.
‘. Data multiplexers, or data selectors
“as they are sometimes known,
generally have one output and several
inputs. Any one of the inputs can, by
placing appropriate logic levels on its
‘control inputs, be routed to its output.
Pata multiplexers thus provide us with
‘a means of sending several different
digital signals along a common signal
dine,
In essence the data multiplexer acts
as a multi-way switch however, by vir-
tue of its internal logic and unlike its
conventional analogue counterpart, the
‘device will only operate with digital
‘signals.
The switch equivalent of the sim-
plest form of data multiplexer is shown
in Fig, 5.1. This two-way arrangement
SELECT] OUTPUT
A
8
‘ SELECT
Fig. 5.1. Simplified switch equivalent
‘of a two-way data multiplexer
r— 4
I
! OUTPUT
i)
a
‘is equivalent to a single-pole double
throw (SPDT) logic switch. The two
‘switch states are controlled by means
of a third select input. When a logic 0
appears on the SELECT input the
switch moves to position A whereas,
when_a logic 1 appears on the SELECT
input the switch moves to position B.
- The internal logic of the two-way '
‘data multiplexer is shown in Fig. 5.2.
A
5
Fig. 5.2. Logic arrangement of a two-
way data multiplexer
lets
4 i
Practi¢al Electronics February 1985
Sequential Logic _
- Techniques Part 5
M.TOOLEY BA and D.WHITFIELD MA MSc C Eng MIEE
This simply consists of two two-input
AND gates, a two-input OR gate and
an inverter. The truth table for this
arrangement is given in Table 5.1. As
0
0
0
1
1
)
1
1
Table 5.1. Truth table for simple two-
way data multiplexer
can be seen, whenever the SELECT in-
put is at logic O the output, Y, takes the
state of the A input wheras, when the
SELECT input is at logic 1, the output
takes the state of the B input.
By grouping together the states for
which the output remains unaffected
by one or other of the inputs (we shall,
for obvious reasons, call these the
“don't care” states!}, the truth table of
the two-way multiplexer can be
reduced to that shown in Table 5.2.
x = don't care
Table 5.2. Simplified version of table
* 6.4
DATA INPUTS
B c
xxxxxx Ob
xxx x =OXx-K x
xx =—=OXK Xx
—-Ox-x x x x x xg
x = don't care
Table 5.3. Truth table for a four-way data multiplexer me,
t=
I
This truth table shows rather more
clearly ‘than its predecessor how the -
SELECT input operates; the X's in the
truth table being used to denote the
“don't care” states.
The switch equivalent of a four-way. —
data multiplexer is shown in Fig, 5.3. ,
A "
1 ! ;
|g ( | ‘
;
to ., | output
DATA INPUTS I Y
SQ $1
—-
SELECT INPUTS .
Fig. 5.3. Simplified switch equivaient,
of a four-way data multiplexer ,c #
Here the output, Y, can be connected
to any one of the four data input lines,,
A to D, by means of an appropriate in-
put on the two select lines, SO and $1.
The truth table for the four-way data
multiplexer is shown in Table 5.3. The
corresponding Boolean expressions
are -
Output Select Inputs
Y=A 50.81
Y=B $0.81
Yat so.$1
Y¥=D $o:,S1
We shall now investigate the operation
of a practical four-way data multi-
plexer, the 74LS153. ; aye
SELECTINPUTS |OUTPUT
a444 w
DIS OD
=200-=008%
gg 8
Ms SEQUENTIAL LOGIC 8
THE 74LS153
The 74LS153 contains two four-
way data multiplexers which have
common select inputs. The pin connec-
tions of the 74LS153 are shown in Fig.
5.4. The two halves of the device
Fig. 5.4. Pin connections for the
74LS153
(referred to as A and B) are convenien-
tly brought out to pins on opposite
sides of the package; the A-side using
pins 1 to 7 whilst the B-side uses pins
9 to 15. Supply connections, using the
conventional pins 8 (OV) and 16 (+5¥V),
are common to both halves of the
device,
Each half of the 74LS153 has its
own active low enable, EN, input.
When these inputs are taken to logic 1
the corresponding outputs immediately
go to logic O irrespective of the state of
any of the data (DO to D3) or select (SO
and $1) inputs.
The internal logic of the 74LS153 is
shown in. Fig. 5.5. This clearly shows
how the EN inputs are gated with the
select, inverted select, and data inputs
at each of the four four-input AND
gates on both sides of the device. The
outputs of each set of AND gates are
then combined in a four-input OR gate.
The 74LS153 is effectively nothing
more complex than a two-pole four-
way switch!
The complete truth table for the
74LS153 is shown in Table 5.4. This
truth table is, of course, identical for
each half of the device. When both
select inputs (SO and $1) are at logic O,
the output (Y) reflects the state of the
DO input. With SO at logic 1 and S1 at
logic O, the output takes the state of
the D1 input, and so on.
sooo
----coee
en nxncond
x xX KX KX ROX XK X a
cevoneneef
woneneeeef
x = don't care
SELECT INPUTS DATA INPUTS
The circuit used for our practical in-
vestigation of the 74LS153 data mui-
tiplexer is shown in Fig. 5.6. It should
CLOCK
$2
LoGic a
$3
$4
Fig. 5.6. Practical arrangement used to
demonstrate the action of the 74LS153
Table 5.4. Truth table for the 74LS$153 data multiplexer
Fig. 5.5. Internal logic of the 74LS153
8
Practical Electronics February 1985
Bok
be noted that only one half of the
device is used. In order to provide four
“different data inputs which may be
readily distinguished from one another,
DO is fed from the clock whilst D1 and
D2 are fed from the Logic Tutor’s
momentary push buttons, $1 and S2
respectively. The remaining data input,
D3, is fed from an inverted clock signal
derived from a 7414 inverter. The rele-
vant half of the device is enabled by
hard wiring the EN input to logic O
whilst the two latching Logic Tutor
switches, S3 and S4, are used to deter-
mine the state of the select inputs, S1
and SO respectively.
The 7414 should be inserted into
socket D whilst the 74LS153 should
be inserted into socket E with the usual
orientation convention (pin-1 of each
device to the respective connection
marked on the Logic Tutor PCB) being
observed. The following links are
required:-
D1 toclock
D2 toE3 (data input D3)
D7 toOV {common}
D16 to +5V (supply)
Ei to logic O (active low enable)
E2 toS3 {select input $1)
E4 toS2 (data input D2)
—E5 toSi (data input D1)
E6 to clock (data input DO)
E7 toD1 (D1 indicates the
output)
E8 toOV (common)
E14 toS4 (select input SO)
E16 to+5V (supply)
{A total of 13 links)
The select inputs should initially
both be set to logic O by appropriate
adjustment of S3 and S4. The output
indicator, |.e.d. 01, should then be seen
to flash ‘on’ and ‘off’ in sympathy with
the clock which is connected to the DO
input line.
S4 should’ now be adjusted to
produce a logic 1 on the SO line whilst
$3 remains at logic 0. In this condition
the |.e.d. will stop flashing and become
extinguished. Now depress $1 to
produce a logic 1 on the D1 input. The
|.e.d. will become illuminated whilst $1
is held down and will become ex-
tinguished again when S11 is released.
$3 and S4 should now be adjusted
to produce logic 1 and logic 0 on the
$1 and SO select inputs respectively.
S2 should now be depressed to
produce a logic 1 on the D2 input. The
led. input will become illuminated for
as long as S2 is held down.
Practical Electronics February 1985
Finally, S4 should be adjusted to
produce a logic 1 input whilst S3
remains at logic 1. In this condition D1
should be seen to flash ‘on’ and ‘off’ in
sympathy with the inverted clock. (i.e.
when the clock l.e.d. is ‘on’, the output
l.e.d. is ‘off’, and vice versa). We can
summarise these observations as
shown in Table 5.5.
| sa | 6a | wn
0 DO (CLOCK)
D1 (S1)}
D2 (S2)
03 [CLOCK)
Table 5.5. Outputs provided by the
circuit of Fig. 5.7. (Note: brackets
indicate Logic Tutor functions)
A PRACTICAL APPLICATION OF
THE 74LS153
We shall now turn our attention to a
simple practical application of the
74LS153 four-way data multiplexer.
Let's assume that we wish to provide
digital selection of the output frequen-
cies of a four-stage binary counter. The
four'Q outputs of the binary counter
can be fed to the four data inputs of the
multiplexer whilst the two select inputs
are fed with a two-bit control signal.
The circuit of a suitable arrangement
is shown in Fig. 5.7. 1C1a forms a sim-
ple relaxation oscillator in which the
output frequency is determined by the
time constant, CxR. !C1b forms an in-
verting buffer, the output of which is a
rectangular pulse wave having a duty
cycle of approximately 1:2 and a fre-
quency of approximately 32Hz. This
signal is then fed to the CLOCK input of
1C2, a 7493 four stage binary counter.
Cla
3 7014
ICib
SELECT | 53
OUTPUT
FREQUENCY | cy
EQUENTIAL LOGIC
In order to enable the normal counting,
sequence, the two master reset inputs,
MR1 and MR2, are taken to logic 0
and the four Q outputs then have fre-
quencies of 16Hz, 8Hz, 4Hz°and 2Hz
approximately.
The 7414 and 74LS153 devices
should be left in sockets D and E whilst,
the 7493 should be inserted in socket
B checking, as usual, that pin-1 aligns
with B1. The following links should
then. be made:-
B1 toBi4
B2 toB3
B3 tologicO
BS to+5V (supply)
B10 toE4
B11 toE5
B12 to OV (common)
B13 to&3
B14 to E6
B16 toD4
D1 tologicO (via a 47uF 25V cap)
D1 toDd2 (via a 470 ohm 0-25W}
D2 toD3
D7 toOV (common)
D16 to+5V = {supply)
E1 to logic O
E2 toS3
E7 toD1 {D1 shows o/p freq.)
E8 toOV
E14 toS4
E16 to+5V (supply)
(A total of 20 links and 2 components)
The two select inputs should first be
set to logic O using S3 and S4. The
output indicator (l.e.d. D1) should then
be seen to flash rapidly ‘on’ and ‘off’
with a frequency of approximately
16Hz. The three other possible settings
of S3 and S4 should then be tested.
NEXT MONTH: De-multiplexers and
time domain multiplexing.
LOGIC o
MR1 = MR2
1C2
7493
Q1
Q@ CLOCK T
1c3
76LS153
OUTPUT
TO OF
EN
LOGIC O
Fig. 5.7. Simple application of the 74LS153
57
Mono/Stereo
Chorus & Flanger .
JOHN M.H.BECKER
fae clock signal that causes the delay chips to sample and
transfer their charges from stage to stage is produced by
[C9 (Fig. 6). This is a standard linear voltage controlled
oscillator chip that produces a squarewave output the fre-
quency of which is related to the value of C24, the current
through VRQ, and the voltage present on pin 9. The single
output from ICQ needs to be split into two opposing phases
as required by the delay chips. If a normal phase split were
to be given then the opposing edges of the antiphase square
waves would coincide. This overlap is prone to causing
system noise from the delay chip outputs even though the
TDA1097 is basically a low noise device, capable of a 77dB
signal-to-noise ratio at a 100kHz clock frequency, though
this degrades slightly with lower clock rates. The overlap on
the edges of the clock is eliminated by the flip flop stage
1C10 in conjunction with the NAND gates IC11a—b. C25 and
R83 slow down the mutual triggering of the flip flop and
gates, resulting in a twin phase output having a short delay
between the respective squarewave edges. Oscillograms
Fig. 7a to 7c show the ‘with’ and ‘without’ effect of the
overlap elimination.
Varying the voltage applied to pin 9 of IC9 varies the clock
frequency. For the automatic modulation of the clock a con-
stantly varying voltage is produced by the low speed triangle
wave oscillator around |C8a—b, and having a frequency
governed by the resistance of VR7 and the value of C22.
{Oscillogram Fig. 8.) Decreasing either increases the output
frequency. The modulation can be switched in and out by
S4, and the level varied from nil to full by the depth control
VR8. C23 slightly rounds off the triangle peaks at faster
modulation speeds. The modulating frequency range is con-
trollable between about 50 milliseconds and 30 seconds,
the clock frequency range is between about 12kHz and
100kHz. For a single delay chip the delay time range is thus
PART TWO
about 64ms to 7-68ms, cascading two delays doubles the
delay times. With the modulating oscillator switched out of
circuit the unit can of course be used as a standard reverb or
short-echo unit, though these effects will not be so pronoun-
ced as those obtainable with the September 1984 PE Echo-
Reverb unit. .
POWER SUPPLY
The unit has been designed to operate from two 9 volt
batteries producing +9V/OV/—9V, and drawing between
13mA and 20mA, depending on the clock oscillator rate. 1C2
and |C3 though do not like a total voltage drop across them
in excess of 16V, which also means that controlling voltages
must not exceed this either. The positive voltage delivered to
IC2, 3, 9, 10 and 11 is thus reduced to a more suitable level
by the drop across the resistor R62 in the delay line bias
divider network. The voltage at R62 is within limits with all
i.c.s in circuit, but may rise if any of the said 5 are not in their
sockets when power is applied. |C9~11 will not mind, but
IC2 and 3 may object. The unit may be operated from a
stabilised power supply if preferred. The acceptable range is
from +5V/OV/—5V to +9V/OV/—S9V. If it is necessary to run
from a power supply greater than +9V/OV/—9V then two
voltage regulator devices should be inserted between the
power supply and the unit as shown in Fig. 9. The voltage
drop across the regulators must be greater than 2V, and R62
may be replaced by a link wire.
CONSTRUCTION
The component layouts for both boards are shown in Figs.
10 and 11. The short link wires on the p.c.b.s can be made
from resistor cut-off leads shaped to the correct spacing
with thin nosed pliers. Sockets should be used with all i.c.s.
The wiring diagram for the unit is shown in Fig. 13. Bring the
TO RG2/C19
+
CLOCK TO
IC? &3
1C8 «PING=—VE ‘
1C8-PIN®= +¥E £30 32 C28 Ulli
1C11* PIN 16=4+¥VE 100n 108n 470p
1Ci|-PIN 7=2=-VE
-¥V
-V
PE16276
Fig. 6. Circuit diagram of the Clock Circuit
Practical Electronics February 1985
Fig. 7a. Usual appearance of two
square-waves without overlap
removal
_——~f
Fig. 8. Modulation oscillator
waveform
connecting wires neatly around the edges of the p.c.b.s to
the controls. The clock leads to IC2 and IC3 should be
brought forward past C19, turn left at the front panel, then
along to the small p.c.b., turn right and connect. Do not take
them on what appears to be the shorter route across the
main p.c.b. as this would direct them across some parts that
might pick up any stray radiation signal. Unless you have the
RITZ ©
i+ C9
[830]
Ie Co)
[Rae
re MEER
c. C12
ao
oS
he leg
ul
«Ty o——~
cae
Fig. 7b. Close up of overlap
OPTIONAL REGULATOR
FOR REDUCING VOLTAGE
FROM EXISTING PSU
TO LEVEL SUITABLE
FOR CHORUS UNIT
PE1633¢
MUSIC PROJECT
Fig. 7c. Accentuated overlap
removal as used in unit
Fig. 7. Clock edge overlap of two anti-phase square-waves
TO UNIT
Fig. 9. Optional regulator circuit
eyes of an eagle, thoroughly check all the p.c.b. joins with a
magnifying glass. Only after all checking has been done
should the i.c.s be inserted into their sockets, remembering
that 1C2, 3, 7,9, 10 and 11 are MOS devices and require the
normal handling precautions. The main point being to keep
yourself and equipment free of static electricity by touching
a grounded source before handling them.
. Fig. 10. Component layout of the Main Board
60
Practical Electronics February 1985
19 20 21 22 23 26 25 SETTING UP
- This is quite straightforward and specialised equipment is
not needed. First, VR1 to VR3 midway, VR4 max resistance
(anticlockwise), VR5 and VR6 min, VR7 to VR9 max, $1 to
S4 off. Plug in a music signal from a prerecorded source into
the X1 socket. Check that the output level reaching the main
amplifier used is the same as the original, Switch on $3
A R77 AICS) enabling the VCA, and bringing up VR6 a change in am-
; plitude and tonal quality should increase. Rotating the clock
oscillator speed control VRQ to its maximum resistance will
slow down the delay and emphasise the double tracking
{cs fs O26 effect. This will be more apparent with staccato sounds
ran ca, rather than mellow drawn out notes. Adjust VR3 around its
midway point until minimum waveform distortion is heard,
which will also coincide with the best delay effect. If an os-
cilloscope is available, the waveform balance will be obvious
when monitoring the output at VR1 and VR2 in the presence
of a strong input signal. {Oscillograms Figs. 12a and 12b,)
Switch on S4 bringing in the sweep
modulator. Varying VR8 will vary the
modulation depth, and VR7 will vary
the modulation rate. Switch off S4,
reset VRQ to slowest clock speed, VR6
to maximum level, switch on S2 for
feedback enabling. Slowly bring up
VRS and a hollowness to the signal
should come in. Maximise VR5 and
carefully reduce the resistance of VR4
Fig. 12a. Sine-wave with VR3 un- Fig. 12b. Sine-wave with both VR1 until the circuit almost goes into full
balanced but VR1 correct and VR3 correct feedback howl. If howl! occurs, sharply
26 27 28 29
Fig. 12. Traces seen at the wiper of VR1
15 27 29 28 21 20
H FRONT PANEL 4
4 |
|
PLP PLEL
|
|
BATT
ON/OFF FEEDBACK DECAY
PCB PIN 12 NOT USED
LINK PCB PIN 13 B22
LINK PCE PIN 16 &25
LINK PCH PIN 17 &24 3
LINK PCB PIN 18 &23 i
Fig. 13. Wiring diagram << i
ae SIND AM Spee ee ee =i)
NOTE R18 R28 ON JACK SOCKETS
4
Practical Electronics February 1985 61
Fig. 14a, Clock residual with VR1
unbalanced (no signal)
Fig. 14b. Clock residual with VR1
balanced (no signal)
Fig. 14c¢. Sine-wave signal with VR1
unbalanced but VR3 correct
Fig. 14. Traces seen at the wiper of VR1
reduce VR4 and start again. Aim for the closest possible to
the howl point. Howl is more likely to occur with strong bass
notes. Switch on $1 to couple the two delay circuits in series
and so produce the double emphasis. If necessary back off
VR4 slightly as the increase in level may kick the circuit into
howl again. If an oscilloscope is not available VR1 and VR2
should be left midway and ignored, otherwise adjust them
for the best balance point of the residual clock frequency in
the absence of an input signal. {Oscillograms Figs. 14a to
14c.) Switch on S4 and experiment with the various settings
until familiar with the control options available, if necessary
readjusting VR3 or VR4.
USE
There are no restrictions to the type of signal fed in
provided that the amplitude is less than the distortion level,
and that the type of music lends itself to enhancement
within the factors discussed earlier and summarised below.
It will soon become obvious which type of music requires
which particular control setting for the best effect. This is a
matter of personal preference, but the author feels that as
with any effects unit, moderation is the keyword. Certainly
overemphasis of an effect is dramatic, but it is easier to
become tired of an over dramatic effect than one which
produces a discrete change. In general terms music having a
high harmonic content, but otherwise of a simple nature, will
benefit most. Mellow or full orchestral sounds will not show
the same degree of change. In the first case there is insuf-
ficient harmonic information available in the signal for the ef-
fect to fully develop. In the second case, the sound is already
so full that the effect will probably be lost amongst the tonal
complexity unless the original sound is full of spiky
waveforms. The harsher sounds of voice, drums, harmonic-
ally rich synthesiser and organ music produce excellent
effects as the waveforms involved are complex. Pure sine
tones and muted waveforms, especially in the lower octaves,
will be less apparent. For the chorusing effect a slower clock
oscillator speed is preferable as the delay time is greater, for
flanging, faster clock speeds are better as the phase shift
occurs then at a more marked rate and spacing.
Photograph illustrating the internal details of the Chorus and Flanger Unit
62
Practical Electronics. February 1985
2.
Ps
ECONOMIC DIGITISER
Sir—Here is a low cost, easy to make project
for use with an X/Y plotter, that can be con-
nected to the expansion interface of the UK
101, or virtually any other micro.
The system consists of a flexible arm,
equipped with two linear potentiometers and a
pointer. This pointer moves over a limited sur-
face of 20 x 20cm. The voltages, generated by
the movement are a measure of the position of
the pointer. Those potentials cannot be used
directly for X/Y plotting, because of their non-
linear output.
A solution of that problem is found, by the
application of the formula, which gives the
relationship between X, Y and the angles of
rotation V and H.
If now the analogue values of these angles
are translated to digital, the calculation with a
BASIC program becomes possible.
The BASIC program can be perceived as a
limited loop with continuous conversion, so
that the manual movement of the pointer can
be digitally stored in memory. It permits also
direct writing to a high resolution display.
Potentiometer
VREEV)
Wooden board
ae
Pointer
Table 1 shows a typical BASIC program
for illustration, and the following notes apply:
5-30 Initialisation of the PIA (A and B) for
input.
VIA (A and B) for output.
40-180 Loop for storing data in memory.
190-280 Loop for playback to X/Y
recorder.
If the digitiser is used with the UK 101, it
may be used with the PE-Expansion Interface
published in Jan—July 1981.
1 do not use the internal A/D and D/A con-
verters, because | wanted a binary indication
for the D/A output and also there is only one
output in the interface. You will find the
schematics joined.
The potentiometers must be of good quality
and very linear. A resolution of approximately
2mm is possible on a surface of 20 x 20cm.
During storing in memory, the values of X
and Y are continually displayed, and must be
integer, positive and between 0 and 255.
The value of L (line 100) is a multiplication
constant to certify an optimum sweep between
O and 255.
The number 255 (line 110) is added to in-
vert the value of Y, which is negative in this
area.
Line 50 in the program is the number of
points you wish to fix. The movement of the
pointer shall not exceed 3cm/sec.
Line 255 defines the speed of playback, and
Prototype digitiser
Limited area
approx. 2S50x 250mm —,
Practical Electronics February 1985
400mm : : creat _V
RO-E
and MICROPROMPT
110
120
130
140
145
150
160
180
190
195
200
210
220
230
240
250
255
260
270
280
REM ; X/Y PLOTTING ROUTINE
P=61340:Q=61342
U=61344:W=61345
POKEU+2,255:POKEW 42,255
POKEP+1,0:POKEP,0:POK EP + 1,255
POKEQ+ 1,0:POKEQ,0:POKEQ +1,
255
INPUT “READY FOR START”:A$
R=5000:S=6000
FOR N=1 TO 200
POKEP+1,60:POKEP + 1,52
V=PEEK(P)/100
POKEQ+1,60:POKEQ + 1,52
H=PEEK(Q)/100
L=150
Y=L*COS(V)+L*COS(V+H):YS
INTCY)+255
X=L*SIN(V)+L*SIN(H+V):X=
INT(X)
PRINT X.Y
POKER,X:POKES,Y
POKEU,X:POKEW,Y
R=R+1:S=S$+1
NEXT N
PRINT“END OF LOOP”
INPUT“DO YOU WANT A
PLAYBACK”;B$
IF BS="Y¥" THEN 210
IF BS="N" THEN 40
R=5000:S=6000
FOR N=! TO 200
X=PEEK(R):Y=PEEK(S)
POKEU,X:POKEW,Y
R=R+1:S=S+1
FOR D=!1 TO 50:NEXT
NEXT N
PRINT“END OF PLAYBACK"
GOTO 190
Table 1. Suggested
software for calculating
the angular co-ordinates
Fig. 1. Mechanical construc-
tion of the digitiser
- Fig. 2. Sample of handwriting
traced and stored from the
digitiser
Wy “
co
65
ICRO-BUS
depends on the type of X/Y recorder (graphic,
scope, screen).
J. Ockier,
Belgium.
DO}
A/D
CONVERTER i Pia
fORY ORT A
p7)
(STARTING
CA PULSE)
oo
‘ Ao ia PIA
CONVERTER PORT B
FOR H 07}
(STARTING
CA2” PULSE)
Fig. 4. Block diagram of the
digitiser
FEY TO DIGITAL INPUTS FROM PIA Nc
a PORT Ass PORTE=Y
GA INAIB!
th
OFFSET
{Tero valte
i Zero ut!
10 CA? FGR
STAATPULS ANALOGUE INPUT WH
9-5 VOLT
O- 255v0LF
- Fig. 5. The ADC channel
DO oO 02 D3 O4
BX 20k
20k
FROM DIGITAL OUTPUTS VIA PORT A=X
007]
Pi D/A FROM VIA
CONVERTER PORT A
07)
TO K/¥
RECORDER
00)
j D/A FROM VIA
‘ CONVERTER PORT &
07 J
\ |
-l\
'
~
' \
' POINTER —
'
'
\ ¥ = L isin + sin( ¥+H)]
1 [FERED] Y= Licos¥ +cos{ ¥+H)]
Fig. 6. A plan view showing the
geometrics of the digitiser. Good
quality highly linear poten-
tiometers must be uséd to obtain
a resolution of around 2mm
IC2
c04042
D6
CZ
OUTPUT aq
SMOOTHING
CIRCUIT
105
Fig. 7. Converting back to analogue for driving a plotter
* BAKER *
GROUP PA. DISCO
AMPLIFERS post £2
150 watt Output, 4 input Mixer pre-amp, Ilustrated fa 3
150 watt Output, Slave 500 mv, input 3 Speaker Outputs. . €80
150+ 150 watt Stereo, 300 watt Mono Slave 500 my. Inputs £125
150 watt P_A. Vocal, 8 inputs. High/Low Mixer Echo Socket £129
100 watt Valve Model, 4 inputs, 5 Outputs. Heavy duty €125
60 watt Mobile 240v AC and 12v DC. 4-68-16 ohn 4 100v line £89
Reverb Unit for Microphone o: Guitar £35 PP £1
Electranic Echo Machine for mic’guitar £86 PP £1. Deluxe £95.
000 0™
BAKER LOUDSPEAKERS Post £2 each
Type Model Size WattsOhms Price
P.A/Disca'Graup OGso'10 10 50 B16 £18.00
Mideange Mid 100/10 10 100 «8 £25.00
Hi-Fi Major 12in 30 84/816 £16.00
Hi-Fi Superb 12in 30 816 += £26.00
P.AJDiscovGroup DG45 12in 45 B16 £16.00
Hi-Fi Woofer 12in 80 8 £25.00
Hi-Fi Auditorium 15m 60 @16 £37.00
P.A/Disca/Group OG75 ts iin 75 W816 £20.00
P.A./Disca'Group 0G100 q2in 100 816 £26.00
PA. Disco’Group 0G10015 Vin 100 B16 £35.00
DISCO CONSOLE Twin Decks, mixer pre amp £145, Carr £10.
Ditto Powered 120 watt £199; of Camplete Disco £300.
120 watt £300; 150 watt £360; 300 watt £410. Carr £30.
DELUXE STEREO DISCO MIXER/EQUALUISER as above plus L.E.D,
VU. displays 5 band graphic equaliser, left/right fader, switchable
inputs for phonesline, miketine
Headphone Monitor, Mike Talkover Switch £1 24 PP £2
As above with 7 Band Graphic £138.
P_A. CABINETS jempty! Single 12 €32; Ooubte 12 £38, carr £10.
WITH SPEAKERS 75W £54; S90W €73; 150W £82; 200W £90.
HORNBOXES 200 Watt £32, 300 Wart £38, Post fa.
WATERPROOF HOANS 8 ohms, 25 watt £22. 30 watt £25. 40 watt
£33. 20W ptus 100 volt line £38. Post €2
MOTOROLA PIEZO ELECTRONIC HORN TWEETER, 334in. square £6
100 watts. No crossover required. 4-8-16 ohm, Fax Main, " £10
CROSSOVERS. TWO-WAY 3000 c’s 30 watt £3. 60 watt 65. 100 wart £6.
3 way 950 cps’3000 cps. 40 watt rating. £4, 60 watt £6.50, 100 watt £10,
LOUDSPEAKER BARGAINS Pitase enquire. many others mt stock,
4 obm, Sen. 7x 4in, £2.50; in, 8xSin. £3, Bin, £3.50. 6\2in 25 £7.50.
Bohm, 2M, Ben, £2; 5x Bin, 6x Ain, 7% 4in, Sin, €2,50; 62m, B x Sin £3; Bin
£4.50; 10in £5; 12in. £6. Bin 25WV £5.50, Bin hi-fi BOW £1250.
15 ohm, 2'un, 31/2in, $x 3in, 6x din, £2.50, 6!in GW £5, Bin. £4. YOn. EF
Z ohm, iq £2: 5 <n, 6 x4in, 7 4in, £250. 120 obm, 3!un dia. £1,
Make Model Size WattsOhms Pnce Post
AUDAX WOOFER slain25 8 e058 £1
GOODMANS HIFAX Tax4v4ain 100 8 Bo &
GOGOMANS HE WOOFER Bin, BB £1350 £1
WHARFEDALE WOOFER bo 30 B 1950 £2
CELESTION DISCO:Group 10in. 50 BG £20 ED
AKAI ‘WOOFER l2in. 808 £16 £2
GOCDMANS HPG/GROUP 12n. 120 @15 EO £2
GOODMANS HPO-DISCO tan 120 WIS €3000 £2
HNH OISCO/GROUP Ton WB16 EM ta
S000MANS HP/BASS Von. 250 8 fers e4
GOODMANS HPD'BASS Min 230 8 iP] i]
METAL GRILLES Bin. £3.00, 10in, £3.50, 12in, £4.50, 15in. £5.50,
18in, £7.50. ee
RCS SOUND TO LIGHT CONTROLLER KIT
Printed circuit. Cabinet. 3 channels 1,000 watts each, Will operate
from Hi-Fi of Disco. £19. Post £t
Ready Built Deluxe 4 Channel 4,000 watt sound chaser + speed +
Programme controls £69. Mk.2 16 programmes, £89, PP £2.
MAINS TRANSFORMERS Price Post
250-0-250V BOMA. £.3V 3.5A. 6.3V 1A £7.00 £2
350-0-350¥ 250m<A. 6.3V 6A CT £12.00 Shrouded £14.00 £2
250V GOmA. 6.3V 2A £4.75 £1
220V 25mA, BV 1 Amp £3.00 220V 45mA_ 6V 2 Amp £4.00 £1
Low voltage tapped outputs available
lamp 6, 8, 10, 12, 16, 1B, 20, 24, 30, 36, 40, 48,60 £6.00 £2
ditto 2 arnp £10.50 3 amp £12.50 5 amp £16.00 £2
31-26-0-26-33 volt 6 amp £14.00 £2
LOW VOLTAGE MAINS TRANSFORMERS £5.50 each post paid
SV, 3A; 12V. 3A; 16V, 2A; 20V, 1A; 30V, 1'2A; 3OV, SA+ 17-0-17V,
2A; 35V. 2A; 20-40-60V, 1A: 12-0.12V, 2A; 20-0-20V, 1A; 50V, 2A.
£8.50 post 50p MINI-MULTI TESTER
Pocket size instrument. o p.v. DC volts 5,
25, 250, 500. AC volts 10, 50, 500, 1000, DC
0-250pa; 0-250ma. Resistance 0 to 600K.
De-Luxe Range Doubler Meter, 50,000
opyv.7» 5 x 2in Resistance 0/20 meg in
5 ranges. Current 50pA to 10A, Volts 0.25"
1000v DC, 10v/1000v AC. £25.00 post £1
PANEL METERS 50}A, 100114, SO0HA, ImA, 5mA, 100MA, 5OOmA,
Tamp, 2 amp, 5 amp, 25 volt, VU 2\4x2«1\4in, £5.50 post 50p
EQUIPMENT CASES. Black Viny! Covered Top, Ali Base
4x 212 « 2\ain. £2.50; 6 » 4 « 1'2in, £3.60; 8 * 5 x Zin, £4.00;
11_* 6 x 3in. £5.50; 1144 x 6 x Sin. £9.00; 15 x 8 x 4in. £12.00.
ALI ANGLE BRACKET 6 = 44 = Jin, 30p.
ALUMINIUM PANELS 18 $.w.g, 12 « 12in. £1.80; 14 * Sin. £1.75;
6 & 4in, 55p; 12 * Bin. £1,30; 1D = Tin. 96p; B x bin. Mp; 14 x 3in,
72p; 12 * Sin, 9p; 16 = 10in, £2.10; 16 = Gin. £1.30,
ALUMINIUM BOXES. MANY OTHER SIZES IN STOCK.
4x 212 2in, £1.20; 3 « 2 x 1in, £1; 6 x 4 ™ 2in. £1,90;8 «6 «3
in. £3.00; 12 x 5 x Jin, £3.60; 6 « 4 =< Jin. £2.20; 10 x 7 « Bin,
£3.60.
HIGH VOLTAGE ELECTROLYTICS
To X/¥ recorder
or oscilloscope
32+ 32/500 £2
20/500 76p 8+ a/500V £1 324.32/350V 75p
3z/350V 45p 8+ 16/450V. 75 32+324+32/450V . £1.50
32/500V Sp 16+16/350V 7p 16+32+32/500V &2
SINGLE PLAY DECKS. Post £2
Make Drive Modei Cartridge Price
GARRARD Rim 6200 Ceram E22
BSR Bett 12Volt Ceramic €20
BSR Belt = P232 Magnetic £28
BSR Rim 207 Ceramic = £20
AUTOCHANGER 8SA Ceramic £20
AUTOCHANGER GARRARD Ceramic £24
DECCA TEAK VENEERED PLINTH space for small amplifier
Board cut for BSR or Garrard 18%4in. x 14\4in. x din. £5. Post £1
TINTED PLASTIC COVERS for Decks. £5 each,
1728 x 1313 x 3\4in, 1814 «122 x Bin. 2112 x 14a x
17« 1248 x Zin. ax I3x 34 142 x 1318 x 24in.
225% x 1378 x 3in. 1698x133 x ain. 21 x 1338 x 41in.
RADIO COMPONENT SPECIALISTS
Dept 4, 337, WHITEHORSE ROAD, CROYDON
SURREY, U.K. Tel: 01-684 1665
Post €1
alain.
| es JUST had an idea. It doesn’t happen
all that often. So when it does I like to tell
somebody about it.
I've been thinking that as we become more
and more a button-pushing civilisation, our
ability to communicate with each other, in
the way we have been doing since the dawn of
mankind, is going to take a nasty knock.
Here’s an extreme instance. If and when
we reach the stage where we wish our
workmates good morning by punching up the
salutation on a keyboard—even though we
may be no more than an office apart—then
sooner or later there can only be one conse-
quence: total atrophy of the vocal chords.
There’s a lot in the old saying, if you don’t
use it you'll lose it.
And think what that could lead to in the
years after you and I are laid in earth.
‘The end of live theatre as we know it. The
death of opera. he finish of slanging
matches in the House of Commons, The
demise of spicy revelations at posh cocktail
parties. Wholesale redundancy of bingo
callers. No more air-cleaning rows between
married couples. No more whispered ex-
changes of sweet nothings between young
lovers. The start of mute TV and the passing
of radio. ‘he amputation of vocal links with
people of other lands... Try that little lot for
starters. Of course, cynics will say that out of
such evil there comes much good.
No more screeching prima donnas. The
end of pompous party political gas. The boon
of being able to get happily stoned at social
functions without having to endure a load of
inconsequential chitchat. No more having to
bear—though Mum might be upset—the
twice-weekly yap of Coronation Street. No
more having to study impossible foreign
languages. Enhanced matrimonial bliss, made
possible by the blessing of mutual silence.
Goodbye, the saints be praised, to Russel
Harty and Michael Parkinson; and a merciful
deliverance from all that is worst in imported
American TV.
x wk & *
‘There's another aspect worth considering, ©
now that we’re diving at an ever-increasing
pace into the electronic age. How is the
keyboard syndrome going to affect the way
we educate our children? What, for example,
is to be the fate of the good old three Rs?
If you can bear a moment of near-geriatric
nostalgia, let me recall Miss Richardson. She
was a grey-haired lady who was totally
dedicated to thumping the rudiments of good
English into the skulls of her elementary
school charges. She had no time for anyone
unwilling to share her love of the glories of
language and gems of literature. We went
Practical Electronics February 1985
V.T.'s views and opinions are entirely his own and not necessarily those of PE
through hell with the old girl. But we
emerged with enriched minds. ‘
Then there was old Bandy Andrews. I take
off my hat to him as well. He firmly believed
that the only worthwhile subjects in the
curriculum were simple addition, division,
multiplication and subtraction. These were
the intellectual vitamins on which he thrived.
Anyone who resisted the same diet was
beyond the pale.
Finally there was Charlie Atkinson—long
since departed for that great big college of
calligraphy in the sky. It was he who trained
us to express our callow thoughts in
splendidly-rounded hands that were a joy to
behold. It was his good luck to pass on before
the ballpoint pen-—-which some people feel
killed individual style stone dead—came into
universal use.
CLES SCSeCeK eee Se wee
"We must all accept, in-
deed embrace, progress
as our sires did the
wheel.”’
ee
I agree that our education in those days
was pretty elementary. But it had soul and
substance. It nurtured the development of
latent talents and equipped us for the years
ahead. Some of us even got places on the
strength of it.
x wk wk *
Leaving aside the emotional ramblings,
let’s look at things in perspective.
The computer, the microprocessor and all
their derivatives, with us and yet to come,
cannot fail to play an enormous part in our
future lives. Nobody but an idiot would deny
that. We must all accept, indeed embrace,
progress as our Sires did the wheel.
But I can’t help feeling that we still need
the basic support of established educational
practices as a preliminary to hurling ourselves
headlong into the fresh technologies. There is
no substitute for a grounding in the three Rs.
Nothing can replace the human larynx as a
channel for human understanding.
Someone is bound to say that writing as I
do for an electronics journal ] ought to be
more aware of which side my wafer is dif-
fused. Point taken, As I said at the beginning,
it’s just an idea. But J still reckon that Miss
Richardson, Bandy Andrews and Charlie
Atkinson still have a job to do.
xk kek
We're all TV critics. Whenever we moan
from our armchairs we're carrying out the
function, even though it’s without an
audience or a reward.
I'm no exception. ‘he other night ] was
watching a programme called “It'll Be
Alright On ‘The Night’, Presented by that
brilliant jester, Denis Norden, it purports to
be a collection of rejected sequences resulting
from cockups by distinguished performers
while recording their programmes.
It is passable entertainment and
moderately funny. And, I suppose, acceptable
to the gullible. But, because I have that kind
of mind, I suspect that some of these boobs
are specifically produced in order to provide a
relatively inexpensive spin-off. Alright, such
conning is perfectly legitimate if it keeps
people happy and laughing.
On the other hand, if these lapses from
professional standards, which betray a rather
irresponsible approach to the job, are
genuine, then ought we not to be just a litte
concerned?
Television is a voracious consumer of time,
talent and, above all, money. Money, by the
way, which you and I help to provide by pass-
ing our crisp oncers across the Post Office
counter every year.
Thinking along those lines, it’s not easy to
accept as a matter of mirth—perhaps even af-
fectionate sympathy—the banalities which
such programmes offer.
Sorry Denis. I’m sure it’s not your fault.
x wk wk *
According to a recent newspaper report,
the robots of the future are going to be a lot
more cuddly. Apparently ‘not tonight darling’
will not be a feature of their synthesised
vocabulary. Sounds promising.
A spokesman for Cardiff University claims
that whispering words of love and affec-
tionate snuggling-up will present no problems
for these romantic devices.
Moreover, he promises us, they will be en-
dowded with limitless energy, Sounds even
more promising.
But will there be anything in the circuitry
to handle that well-known limitation—the
headache?
x &k& k
In the meantime, it is reported that
America has added a new category to census
statistics: —Robots.
The first robot count is underway this
month and will be aimed at robots on factory
assembly lines, The special census form also
has a section to collect information about
home robots.
A more frightening rumour is the news
that special robots are being seriously con-
sidered for duties as “personal” and home
security guards. These robots are, it is
claimed, programmed to deliver varying
degrees of bodily harm; from simple electric
shocks to “dismemberment”.
It appears that the only obstacle is the ex-
pected multi-million dollar lawsuits that
could be lodged by injured parties,—A case
for Robot against Robot?
67
PRACTICAL ELECTRONICS sy
PRINTED CIRCUIT BOARD SERVICE —
Printed circuit boards for certain PE constructional projects are now
available from the PE PCB Service, see list. They are fully drilled
and roller tinned. All prices include VAT and. postage and packing.
Add £1 per board for overseas airmail. Remittances should be sent
to: PE PCB Service, Practical Electronics Editorial Offices,
Westover House, West Quay Road, Poole, Dorset BH15
1JG. Cheques should be crossed and made payable to IPC
Magazines Ltd.
Please note that when ordering it is important to give project
title, order code and the quantity. Please print name and address in
Block Caps. Do not send any other correspondence with your order.
Readers are advised to check with prices appearing in the
current issue before ordering.
NOTE: Please allow 28 days for delivery. We can only
supply boards listed here or in the November 1984 issue.
PROJECT TITLE
FEB ‘81
Slave Light Dimmer
MAR ‘81
27/28MHz Converter
Microphone Mixer
Period Power Tester
APRIL '81
Speech Processor
Mini Drill
103-01 | £1.79
103-02 | £1.83
103-03 | £2.25
104-01 | £1.55
104-02 | £1.50
105-01 | £6.65
105-02 | £1.49
Digisounder
Thermometer
Horologicum
Analogue Frequency Meter
Ignition System
APRIL ‘82
Med, Resolution Equaliser (UK 101)
Enlarger Timer
AU 8
Automatic Photographer
Home Alarm
Into the Real World
Accessory PSU
44 Digit Frequency Meter
APRIL '83
Phaser
JUNE ‘8
| Program Conditioner 306-01
SEPT ‘83
Guitar Active Tone Control!
Ground Communication System
PROJECT TITLE
MAR ‘84
Spectrum Autosave
Microstepper
Sustain Unit
Audio Signal Generator
Simple Logic Generator
EPROM Duplicator
Alarm System
Oscilloscope Calibrator
AUG '84
Comm, 64 RS232C Interface
Field Measurement
Digital Dice
Simple Logic Analyser
Alarm System
SEPT ‘84
Parallel to Serial Converter
Through the Mains Controller
O "84
Logic Probe 410-01
NOW ‘84
Computer OFM Adaptor 411-01
DEC 84
Ni-Cad Charger
JAN ‘85
Outrider Car Computer (Set of 2 boards}
EB ‘85
Modular Audio Power System
Pt-1: Power Amp Board
Spectrum DAC/ADC Board
eee ee ee ee eee ee ee ee
i PE PRINTED CIRCUIT BOARD SERVICE |
I Please send me the following p.c.b-s. |
w Order Code Quantity Price
nw
| hi iepe fewasetee: vecalnienterdy euepeecmaases |
De Saga ivasiamna> sien apie ote ebe a8 semgucrrnmend |
7)
| P| ike KES AON RAGE yblpewert oa bawedul wu Raed eee Ba ee |
| EEO HS eee eta a eS edema iote weeds &) oe eWe Sas, © O Rae |
| $ | enclose cheque /PO for £ o.........cc. ccc cece cee s eee ees |
| A PSU aos as ea agers Sad nSas rome aesn ae erat a dus eau auk |
| a Address, a: seney eivia uy ad eee evan eae ee ada, Subsea bs |
a
|
L
Practical Electronics February 1985
From a gentle purr to a mighty roar,
the tightly controlled power of the
beast is yours to command!
A new range of superb quality loudspeakers.
* Virtually indestructible high temperature
voice-coil reinforced with glass-fibre
* 100% heat overload tolerance
% Advanced technology magnet system
* Rigid cast alloy chassis
* Linen or Plastiflex elastomer surrounds
* 5-year guarantee (in addition to statutory rights)
Available in 5, 8, 10, 12, 15 and 18 inch models with 82 and some 162.
impedances and with input powers ranging from 50W to 300W e.g.
Sin. 50W 95dB 80): XG39N / 160: XG40T £17.95§
Bin. 100W 98dB 81): XG43W £29.95§
10in. 100W 100dB 811: XG46A £29.95§
12in. 100W 101dB 81): XG49D £29.95§
12in. Twin Cone 100W 100dB 81): XGS5OE / 162: XG51F £31.95§
Note - the output power doubles for each 30B increase (ref 1W @ 1m).
A new range of very high quality multimeters offering truly amazing
quality at the price.
Pocket Multimeter, 16 ranges, 200001/V DC/AC £6.95§ (YJO6G)
M-102BZ with Continuity buzzer, battery tester and 10A DC range, 23 ranges,
20,00012/V DC £14.95§ (YJO7H)
M-2020S with Transistor, Diode & LED tester and 10A DC range, 27 ranges
20,00002/V DC £19.95§ (YJO8J)
M-5050E Electronic Multimeter with very high impedance, FET input, 53
ranges including peak-to-peak AC. centre-zero and 12A AC/DC ranges
£34.95§ (YJO9K)
M-5010 Digital Multimeter with 31 ranges including 209 and 20nA DC/AC FSD
ranges, continuity buzzer, diode test, and gold-plated PCB for long-term
reliability and consistent high accuracy (0.25% +1 digit DCV) £42.50§ (YJ10L)
N.B. All our prices include VAT and Carriage. A 50p handling charge must be
added if your total order is less than £5 on mail order (except catalogue)
Mail Order: P.O. Box 3, Rayleigh, Essex SS6 8LR. Tel: Southend (0702) 552911
SHOPS
® BIRMINGHAM Lynton Square, Perry Barr, Tel: 021-356 7292.
® LONDON 159-161 King Street, Hammersmith, W6. Tel: 01-748 0926.
® MANCHESTER 8 Oxford Road, Tel: 061-236 0281.
© SOUTHAMPTON 46-48 Bevois Valley Road, Tel: 0703 25831.
® SOUTHEND 282-284 London Rd, Westcliff-on-Sea, Essex. Tel: 0702-554000
Shops closed all day Monday.
§ Indicates that a lower price is available in our shops.
Lhe y/)
| Our huge range of top quality electronic components at very competitive
| prices are all detailed in our catalogue, and with well over 600 new lines
in our 1985 edition and many design improvements, it’s well worth
getting a copy. Here are just a few examples from the catalogue.
(The items below are NOT kits).
* Most phono and jack plugs now with integral strain relief sleeve - gold-plated
types also available from 14p (gold from 70p)
* Stereo Disco Mixer with cross-fade, talk-over, cue monitoring, aux input,
slide controls. Only £58.95 (AF99H)
* 10-Channel Stereo Graphic Equalisers - 3 models - basic; with
meter; and with spectrum analyser - from £77.95
* Digital Delay Line permits Slap-back, Doubling, Flanging, Chorus and Echo,
11 controls. Only £195.00 (AF98G)
* Video Enhancer improves picture quality when recording from one VTR to
another, and with TV's with monitor input. Only 28.95 (XG59P)
* Detailed descriptions of the exciting new 74HC range of IC's which combine
the advantages of CMOS and TTL. From 46p
* Keyboards: sloping keys, two-tone grey, mounted in steel frame, very smart
cases (extra) available. 61 keys, only £33.95 (YJ12N)
79 keys, only £37.95 (YJ13P)
* 1% Resistors now 50ppm/*C, 0.4W, only 2p each!
* Auto transformers 120/240V 50VA, £10.75§ (YJ56L). 100VA £14.95§
(YJ57M). 150VA £16.95§ (YJ58N). 250VA £21.95§ (YJ59P).
.95 (FK51F)
peak level
Pick up a copy now at any branch of W.H.
"= Phone before 2pm for Smith or in one of our shops. The price is
| same prefers g | still just £1.35, or £1.75 by post from our
Bs | Rayleigh address (quote CA02C).
|
L
Post this coupon now for your copy of the 1985 catalogue.
Price £1.35 + 40p post and packing. If you live outside the U.K,
send £2.40 or 11 International Reply Coupons. | enclose £1.75.
All offers subject to availability.
Prices firm until Feb 9th 1985.
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11